Amelogenesis Imperfecta

Class of hereditary disorders of dental enamel development, isolated or syndromic, with more than seventy known genes and four major clinical phenotypes: hypoplastic, hypomaturation, hypocalcified and hypomaturation-hypoplastic with taurodontism

Literature and regulatory information updated: 29 October 2026.

0. Initial Reference Table

Table sources: OMIM (August 2026 morbid map), Orphanet and GeneGuidance [3–5, 18, 59–66].

1. Introduction

1.1. Definition and boundaries of the class

Amelogenesis imperfecta (AI) designates a group of conditions of genomic origin that alter the structure and clinical appearance of the enamel of all or nearly all teeth, more or less equally, and that may be associated with morphological or biochemical changes in other organs [1, 67]. The definition has three practical consequences. The first is that the defect affects both dentitions, although in some forms the primary dentition appears less affected [1, 23]. The second is that the pattern does not follow the chronology of tooth formation, which distinguishes it from hypoplasias caused by time-limited insults, such as severe systemic illness in childhood, antileukaemic therapy or excessive fluoride intake [1, 22]. The third is that AI is not restricted to disorders confined to the teeth: the 2003 definition, taken up by the 2007 Orphanet review, accepts syndromic forms, and the experience of the following two decades, with FAM20A and WDR72, showed that forms previously considered isolated were in fact syndromic [1, 2].

The class has become very heterogeneous. In 2023, the French team of the O-Rares reference centre counted more than 70 genes associated with isolated or syndromic AI, and the ERN CRANIO orodental group refers to more than 100 distinct entities [2, 20]. A survey of 408 rare diseases with orodental manifestations identified 105 with enamel defects, 21 isolated and 84 syndromic [2]. Molar-incisor hypomineralisation (MIH), a common defect of unclear aetiology affecting one or more first permanent molars and sometimes incisors, is not classified as AI [1, 68].

Genetic dental anomalies as a whole (tooth agenesis, dentine disorders, primary failure of eruption) enter the differential diagnosis; this article explores the AI class in depth following the structure used for syndromic classes. Syndromic forms receive here the description needed for recognition starting from the enamel: Zellweger spectrum syndromes, including Heimler syndrome, hereditary renal tubular acidoses, cone and cone-rod dystrophies, including Jalili syndrome, and trichothiodystrophy.

1.2. Historical milestones

AI was described as an entity separate from dentinogenesis imperfecta by Finn in 1938 [6]. Weinmann and colleagues proposed in 1945 the first division, purely phenotypic, between hypoplastic and hypocalcified forms [1]. From 1957, Witkop added hypomaturation and then the mode of inheritance as a delimiting criterion, and his 1988 classification, with four types and fifteen subtypes, remains the most widely used in clinical practice and teaching [1, 2, 23, 69]. In 1995, Aldred and Crawford proposed including the molecular defect and the biochemical result, and in 2003 Aldred, Savarirayan and Crawford argued that inheritance should be the primary criterion and phenotype the secondary one, until molecular diagnosis became available [1, 67].

The molecular era began in 1991 with the identification of a deletion in AMELX (then designated AMG) in a large Swedish family with X-linked AI [6, 70]. This was followed by ENAM (2001–2002), KLK4 (2004), MMP20 (2005), FAM83H (2007–2008), WDR72 (2009), FAM20A (2011), ODAPH/C4orf26 (2012), SLC24A4 (2013), dominant LAMB3, ITGB6 and AMBN (2013–2014), AMTN (2016), GPR68 and ACP4 (2016), RELT (2019) and SP6 (2020) [6, 71–83]. In 2007, the Orphanet review considered genetic diagnosis “only a research tool” [1]; in 2024, a Swedish review concluded that the genetic evidence made it advisable to recommend genetic testing as a complement to clinical diagnosis, and the 2026 German S3 guideline recommends it when the clinical and radiological picture suggests a genetic cause [24, 84].

1.3. Classifications in use

The following table summarises the phenotypic and mixed classifications used in the literature, with the dates and criteria of each [1, 2, 67].

Witkop's 1988 classification, with the gene correspondence proposed in 2023 and the OMIM entries, is the reference shared by clinicians and laboratories [2, 4, 69].

The OMIM nomenclature (AI1A to AI4) derives from Witkop's, but the current entries are assigned by gene and do not fully coincide with the phenotypic subtypes: for example, Witkop type IG designates recessive enamel agenesis, whereas the OMIM entry AI1G corresponds to enamel-renal syndrome due to FAM20A; and the entries AI1H (ITGB6), AI1J (ACP4) and AI1K (SP6) did not exist in the original classification [2, 4]. Orphanet groups the OMIM entries into four clinical subtypes: hypoplastic (ORPHA:100031: 104500, 104530, 204650, 301201, 616221, 616270, 617297), hypocalcified (ORPHA:100032: 130900, 616221, 617607, 618386), hypomaturation (ORPHA:100033: 204700, 301200, 612529, 613211, 614832, 615887, 617217) and hypomaturation-hypoplastic with taurodontism (ORPHA:100034: 104510); ITGB6-related AI appears in two subtypes, which reflects its mixed phenotype [8–11]. The disease codes of the international classifications are poorly discriminating: ICD-9 grouped AI with dentinogenesis imperfecta, dentine dysplasia and “shell teeth” under a single code (520.5), and SNOMED uses a single morphological category [1].

1.4. Definitions of the phenotypes

Hypoplastic AI corresponds to a quantitative defect: the enamel is thin but mineralised, with pits, grooves or striations, and at the extreme it may be absent (enamel agenesis); it results from failure of the secretory stage of amelogenesis [2, 6]. Hypomineralised AI corresponds to failure of the maturation stage, with enamel of full thickness but weak, which fails early; it is subdivided into hypomaturation, due to incomplete removal of matrix protein, with brittle, opaque and sometimes pigmented enamel, and hypocalcified, due to insufficient calcium transport, with soft enamel [6]. The Strasbourg team stressed in 2023 that the terms hypocalcification and hypomaturation were coined before the genes were known, that both phenotypes have hypomineralisation as their primary feature and that, clinically, “hypocalcified” describes a softer enamel with post-eruptive breakdown and “hypomature” a less mineralised enamel that is nonetheless resistant enough to preserve tooth morphology; it therefore proposed merging classes II and III [2]. The 2026 British consensus likewise treats the hypocalcified and hypomature phenotypes as a spectrum of severity of hypomineralisation [23]. The hypomaturation-hypoplastic form with taurodontism combines the two defects with enlarged pulp chambers and root shortening, and forms an allelic spectrum with tricho-dento-osseous syndrome [6, 85].

2. Molecular Genetics and Pathophysiology

2.1. Amelogenesis and the three compartments

Enamel is the most highly mineralised tissue in the body: in the mature state, more than 95% of its volume is occupied by large, rigorously organised hydroxyapatite crystals, and the mineral content reaches 96% to 98% [1, 2]. It is also the only mineralised tissue of ectodermal origin and it does not regenerate, because the ameloblasts that produce it disappear when the tooth erupts; any formation defect is therefore permanently recorded [1, 2]. Amelogenesis proceeds in three stages. In the secretory stage, the ameloblasts move away from the amelodentinal junction and secrete, from Tomes' processes, a protein matrix in which immature crystallites are deposited and grow in length; at the end of this stage, the forming enamel contains about 30% mineral and 25% protein. In the transition, secretion decreases, the ameloblasts halve in height and about 25% die by apoptosis. In maturation, proteases degrade the matrix, the ameloblasts increase active transport of mineral ions and the crystallites grow in width and thickness, alternating cyclically between a ruffle-ended form, which pumps protons and maintains a slightly acidic pH, and a smooth-ended form, with a more neutral pH; at the end, the enamel is almost 100% mineral and about half of the ameloblasts have undergone apoptosis [6].

The 2017 Leeds review organised the proteins involved in AI according to three compartments: the enamel organ (ameloblasts and supporting cells), the extracellular space in which the matrix mineralises, and the interface between them [6]. The proteins are grouped into matrix proteins (amelogenin, enamelin, ameloblastin), matrix proteases (MMP20, KLK4), cell-cell and cell-matrix adhesion proteins (ITGB6, LAMB3, LAMA3, COL17A1, AMTN, FAM83H), transport proteins (WDR72, SLC24A4), master regulators (FAM20A, DLX3) and a group with diverse functions (ODAPH in crystal nucleation, GPR68 as a proton sensor, ACP4 of unknown function) [6]. A cross-cutting lesson is that absence of the protein is not always the mechanism: some Amelx and Enam variants in the mouse activate the unfolded protein response and ameloblast apoptosis, and in several genes (FAM83H, LAMB3, ENAM, ODAPH) pathogenic variants cluster in positions that escape nonsense-mediated messenger RNA decay (NMD), producing truncated proteins with a dominant effect [6].

2.2. Enamel matrix proteins

The matrix proteins (amelogenin, ameloblastin, enamelin and amelotin) derive from a common ancestral gene of the secretory calcium-binding phosphoprotein cluster, have a signal peptide and a casein kinase 2 phosphorylation domain, a probable target of the Golgi kinase FAM20C [6]. Amelogenin accounts for about 90% of the secreted proteins; ameloblastin and enamelin make up the remaining 10% [6].

AMELX. The amelogenin gene (OMIM 300391 [86]; Xp22.2) has seven exons, lies within an intron of ARHGAP6 and encodes a hydrophobic protein, rich in proline and histidine, that acts as a pH buffer for the matrix and as a scaffold for the spacing and growth of crystallites [1, 6, 87]. The gametologue AMELY, on the Y chromosome, is transcribed at about 10% of the level of AMELX and does not compensate for its loss; men with a null AMELY allele, identified in forensic series, are phenotypically normal, and AMELY is not an enamel disease gene [5, 6]. X-linked AI (AI1E, OMIM #301200 [28]) is classified by OMIM as X-linked dominant; the Oxford Desk Reference describes it as X-linked recessive with female carriers who show enamel ridges, which reflects the same reality of more mildly affected women [4, 88]. In males, the phenotype depends on the type and position of the variant: large deletions and N-terminal variants produce hypomaturation AI with variable focal hypoplasia and a “snow-capped” appearance, whereas signal peptide variants and truncations close to the C-terminus produce smooth hypoplastic AI or almost absent enamel, attributed to a toxic effect of the non-secreted mutant protein [5, 6, 89]. In heterozygous females, random X inactivation produces vertical bands of normal and abnormal enamel or vertical grooves and ridges, and the proportion follows the skewing of inactivation [1, 5]. More than twenty variants have been described (large deletions, frameshifts, nonsense codons, missense substitutions); whole-gene deletions make it necessary to look for copy number variation when sequencing is negative [5, 6]. The silent variant NM_182680.1:c.120T>C prevents skipping of exon 4 in most transcripts and causes AI, which shows that alternative splicing is relevant to the disease [6]. One point of nomenclature is critical: the AI LOVD (Leiden Open Variation Database) and the older literature number variants on transcript NM_182680.1, which includes exon 4, whereas the MANE (Matched Annotation from NCBI and EMBL-EBI) Select transcript is NM_001142.2, which is shorter; the recurrent North American variant described as c.208C>A p.(Pro70Thr) corresponds to NM_001142.2:c.166C>A p.(Pro56Thr), and c.473delC p.(Pro158HisfsTer31) corresponds to NM_001142.2:c.431del p.(Pro144HisfsTer31) [6, 87].

ENAM. Enamelin (OMIM 606585 [90]; 4q13.3) is the largest of the matrix proteins, tooth-specific, acidic, expressed mainly in the secretory stage and cleaved successively from the C-terminus; the intact protein is found only in the outermost layer of newly secreted matrix and participates in crystallite elongation [6]. The deficiency is quantitative, with hypoplastic enamel [5]. Allelic dosage determines the phenotype: heterozygotes have AI1B (OMIM #104500 [26]), with localised hypoplasia (horizontal rows of pits, grooves or a band of thin enamel on the buccal surfaces and cuspal thirds) or generalised smooth hypoplasia with small, spaced, yellow teeth; homozygotes and compound heterozygotes have AI1C (OMIM #204650 [27]), with generally thin or absent enamel and anterior open bite [5, 91, 92]. In consanguineous Turkish families, homozygotes for the two-base insertion g.13185_13186insAG had generalised AI with open bite and heterozygotes only localised pits [1, 91]. The first ENAM variant, a splicing variant, produced severe smooth dominant AI through a dominant negative effect, and a nonsense variant segregated as dominant local hypoplasia (AIH2 locus) [6, 71]. Zygosity-dependent severity is more frequent with truncating variants that escape NMD [6]. The recurrent variants, with MANE transcript NM_031889.3, are c.92T>G p.(Leu31Arg), in five British families and possibly a founder variant; c.157A>T p.(Lys53Ter), in seven Swedish families; c.588+1del, in five families of Japanese, Lebanese and Slovenian origin; and c.1259_1260insAG p.(Pro422ValfsTer27), in seven Turkish and Slovenian families — the protein notation p.(V422Pfs*27) in the AI LOVD does not match validation on the MANE transcript [6, 87]. ENAM and AMBN are adjacent genes in the same cluster, and the two entities should not be confused [5].

AMBN. Ameloblastin (OMIM 601259 [93]; 4q13.3) is the second most abundant matrix protein, required for formation of the initial enamel ribbon and for appositional growth; in the Ambn-null mouse, the ameloblasts do not initiate apposition, degenerate and deposit ectopic mineral, and the dentine is normal [5, 6]. Biallelic loss causes AI1F (OMIM #616270 [30]), yellow hypoplastic AI with thin, largely aprismatic enamel: the first family, consanguineous, from Costa Rica, had an in-frame deletion of exon 6 that shortens the protein from 447 to 368 residues [5, 6, 78]. In 2024, a series of 11 patients with 5 novel variants showed three groups: biallelic families with yellow hypoplastic AI, two monoallelic families with hypomaturation AI with near-normal volume and pits, and three monoallelic families with white hypoplastic AI with thin but intact enamel, which documents recessive and dominant inheritance [7]. A family with the heterozygous substitution NM_016519.6:c.1069C>T p.(Pro357Ser) had severe dominant AI with dentine involvement, short roots and pulp obliteration, the only report of AMBN with dentine disease [5]. Orphanet records the inheritance of 616270 as dominant and OMIM as recessive; both exist [5, 7].

AMTN. Amelotin (OMIM 610912 [94]; 4q13.3) is secreted by transition- and maturation-stage ameloblasts, is located in the basal lamina, binds ODAM and SCPPPQ1 and participates in forming the compact aprismatic surface layer; the Amtn-null mouse has chalky-looking incisors and soft surface enamel [6]. The association with AI3B (OMIM #617607 [42]) is provisional and rests on a single family from Costa Rica with dominant hypomineralised AI and a heterozygous 8678 bp deletion that removes exons 3 to 6 (in-frame loss of 92 amino acids); as of September 2026 a second independent family had not been published [5, 79].

2.3. Matrix proteases

MMP20. Enamelysin (OMIM 604629 [95]; 11q22.2) is a zinc-dependent endopeptidase, secreted in inactive form by the ameloblasts at the same time as the matrix proteins, which it cleaves at specific residues shortly after secretion; it controls crystal morphology and cleaves cadherins, allowing coordinated movement of the ameloblasts [6]. Biallelic loss causes pigmented hypomaturation AI (AI2A2, OMIM #612529 [36]), first described in 2005 [73]. Missense (in the catalytic domain or the haemopexin domain), nonsense, frameshift and splicing variants have been described, all with a similar phenotype; active-site substitutions abolish catalysis and heterozygotes are unaffected [5, 6]. In the AI LOVD, MMP20 accounted for 13 families (4.8%), and in the 2023 French cohort it was, together with FAM83H, the most frequent gene in isolated AI [2, 6]. The splicing variant NM_004771.4:c.954-2A>T was found in four North American, French, German and Moroccan families [6, 87].

KLK4. Kallikrein 4 (OMIM 603767 [96]; 19q13.41) is a serine protease secreted in the transition and maturation stages, activated by MMP20 and able to inactivate it, which degrades matrix proteins over a wide pH range so that the ameloblasts can remove them [6]. Biallelic loss causes hypomaturation AI (AI2A1, OMIM #204700 [35]), with enamel of normal thickness, yellow to brown, with an “agar jelly” appearance, in which the inner enamel is less mineralised than the outer [5, 6, 72]. Few families have been described, mostly consanguineous; the variant NM_004917.5:c.632del p.(Leu211ArgfsTer37), with a frequency of 0.15% in the South Asian population, was found in five Pakistani families [6, 87].

2.4. Cell-cell and cell-matrix adhesion

LAMB3, LAMA3 and COL17A1. Laminin-332 (alpha-3, beta-3 and gamma-2 chains, encoded by LAMA3, LAMB3 and LAMC2) and collagen XVII (COL17A1) are components of the hemidesmosomes (LAMB3 gene: OMIM 150310 [97]) that anchor epithelial cells to the basement membrane; in mature ameloblasts and in Tomes' processes, laminin-332 participates in differentiation and in adhesion to the enamel matrix [6]. Biallelic loss of these genes causes junctional epidermolysis bullosa (JEB), autosomal recessive, in which LAMB3 accounts for about 70% of cases, COL17A1 12%, LAMA3 9%, LAMC2 9% and ITGB4 less than 1% [5, 15]. Pitted AI is common in severe generalised JEB and persists as a feature of the intermediate form compatible with adult life, and the teeth of patients with JEB show disorganisation of the ameloblasts and changes in the chemical composition of the enamel [6, 15]. Heterozygous carriers of some LAMA3, LAMB3 and COL17A1 variants have hypoplastic AI without skin fragility (AI1A, OMIM #104530 [25]) [6, 15]. The mechanisms differ between genes: LAMA3 variants (for example, c.488delG) and truncating COL17A1 variants act through haploinsufficiency and cause JEB in the biallelic state, and glycine substitutions in the triple helix are also described in COL17A1; most LAMB3 variants associated with AI, by contrast, are truncating in the penultimate or last exon, escape NMD and are compatible with a dominant gain-of-function effect, with no report in patients with JEB [6]. In two families with dominant AI, severe generalised hypoplasia with deep linear grooves and pits, the exome identified in the last exon of LAMB3 the deletion NM_000228.3:c.3446_3453del p.(Gly1149GlufsTer8) and the nonsense variant c.3431C>A p.(Ser1144Ter) [77, 87]. Two additional families had a nonsense variant in the last exon and a splice acceptor variant in the last intron, with well-mineralised but pitted enamel, and cervical enamel less affected than coronal enamel [5]. The practical consequence is that the position of a truncating LAMB3 variant matters: a heterozygous variant in the last exon in an individual with pitted enamel and normal skin is compatible with dominant AI and does not predict JEB in that individual [5]. The frequency of AI in JEB carriers is probably underestimated, because JEB reports mention the enamel only in passing [6].

ITGB6. The integrin beta-6 subunit (OMIM 147558 [98]; 2q24.2) pairs only with the alpha-V subunit, recognises RGD motifs in fibronectin, tenascin and fibrillin-1 and activates transforming growth factor beta-1; it is expressed mainly in maturation-stage ameloblasts [5, 6]. The Itgb6-null mouse has hypomineralised AI without prismatic arrangement and with pits [6]. In patients (AI1H, OMIM #616221 [32], recessive), the phenotype is mixed: generalised hypoplastic with a rough surface, hypomineralised with pits or hypoplastic-hypomineralised, with reduced microhardness of enamel and dentine, prism collapse and, in one patient, taurodontism; the diagnosis should not be excluded because the enamel does not fit a single Witkop type [5, 6]. Up to 2023 six families and eight variants had been published; in 2026 two further families were described with compound heterozygous variants (NM_000888.5: c.1060C>T p.(Gln354*) with c.2312A>G p.(Asn771Ser); c.1693T>C p.(Cys565Arg) with c.2091delC p.(Asn698Metfs*13)), both with generalised hypoplastic and hypomineralised AI without extraoral symptoms [5, 99]. The proposed expansion to alopecia, intellectual disability and dento-gingival changes rests on a single consanguineous Pakistani family and has not been adopted by OMIM [5, 6].

FAM83H. The FAM83H gene (OMIM 611927 [100]; 8q24.3; current symbol SACK1H) encodes a ubiquitous intracellular protein, with maximal expression in pre-secretory and secretory ameloblasts, which, through binding to casein kinase 1, organises the keratin cytoskeleton and participates in desmosome formation [6]. Heterozygous variants cause dominant hypocalcified AI (AI3A, OMIM #130900 [41]), with prism defects mainly near the amelodentinal junction and increased organic content [6]. By 2026, 39 causal variants had been described, all in the last exon (exon 5), 92% truncating (31 nonsense and 5 frameshift), concentrated in the region encoding Ser287 to Glu694; three missense variants produce milder phenotypes [6, 101]. Truncating variants in the last exon escape NMD and the truncated proteins accumulate in the nucleus; the shortest (p.Ser287*) binds and inhibits casein kinase 1; Fam83h-null or overexpressing mice do not have AI, which supports a dominant negative or gain-of-function effect [6]. More distal truncations are associated with a milder phenotype, confined to the cervical regions [6]. The recurrent variants, which behave as mutational hotspots, include NM_198488.5:c.1192C>T p.(Gln398Ter), c.1289C>T p.(Ser430Ter), c.1354C>T p.(Gln452Ter), c.1379G>A p.(Trp460Ter) (five families), c.1387C>T p.(Gln463Ter) and c.2029C>T p.(Gln677Ter) [87, 101]. The variant c.973C>T corresponds to p.(Arg325Ter); the notation “p.(L31R)” shown for this variant in the AI LOVD table is a transcription error [6, 87]. FAM83H was the most frequent gene in the AI LOVD (52 families, 19.3%) and in a 2011 North American cohort (12 of 26 solved families, 46%) [6, 102]. Fam83h-null mice have delayed eruption, a smaller pulp chamber and impaired mandibular growth, and two Brazilian families described in 2026 had impacted teeth with pre-eruptive coronal resorption in 63% to 82% of the impacted teeth, which suggests a role in eruption and in the dental follicle [101].

2.5. Transport, pH and mineralisation

WDR72. The WDR72 gene (OMIM 613214 [103]; 15q21.3) encodes a vesicle coat protein with eight WD40 repeats arranged in a beta-propeller, most highly expressed in maturation-stage ameloblasts and, outside the tooth, in the bladder and kidney [5, 6]. Wdr72-null mice have hypomaturation AI, shorter maturation ameloblasts, amelogenin retention, absence of ruffle-ended ameloblasts and altered localisation of the calcium transporter SLC24A4 [6]. Biallelic loss causes hypomaturation AI (AI2A3, OMIM #613211 [37]), described in 2009 in families with truncating variants, all subject to NMD and with an identical phenotype [6, 74]. The same vesicle trafficking function is required in the alpha-intercalated cells of the collecting duct: the 2025 GeneReviews includes WDR72 among the genes of autosomal recessive hereditary distal renal tubular acidosis (dRTA), with AI as the extrarenal hallmark, and most patients described with WDR72-related dRTA are of Indian ancestry [14]. The protein binds subunits of the vacuolar-type proton ATPase [14]. Variants between the two beta-propeller clusters are associated with hypodontia and delayed eruption; the short stature reported in consanguineous families has not been shown to be an effect of WDR72 [6]. The variant NM_182758.4:c.1467_1468del p.(Val491AspfsTer8) was described in three families of Mexican and Turkish origin [6, 87].

SLC24A4. The potassium-dependent sodium-calcium exchanger NCKX4 (OMIM 609840 [104]; 14q32.12) exchanges one calcium ion and one potassium ion for four sodium ions; it is overexpressed in ruffle-ended maturation ameloblasts, on the membrane facing the enamel, and transports calcium into the matrix [5, 6]. Biallelic loss causes hypomaturation AI (AI2A5, OMIM #615887 [39]), identified in 2013 by autozygosity mapping and exome sequencing [76]. The mutant proteins reach the membrane but do not transport calcium, or transport 78% less, so the defect is functional and not one of localisation [5]. Common variants at the same locus are associated with normal skin, hair and eye pigmentation, a normal trait that should not be confused with the enamel disease [5].

GPR68. The proton-sensing G protein-coupled receptor OGR1 (OMIM 601404 [105]; 14q32.11) is almost silent at pH 7.8 and fully activated at pH 6.8, exactly the range the matrix passes through during maturation; it is expressed in ameloblasts at all stages and has been proposed as the sensor that drives the alternation between ruffle-ended and smooth-ended forms [5, 6]. Biallelic loss causes hypomaturation AI (AI2A6, OMIM #617217 [40]), described in 2016 in three families with a large in-frame deletion, a frameshift and a missense variant; the variant NM_001177676.2:c.221T>C p.(Leu74Pro) reaches the membrane but loses pH sensitivity [5, 80].

ODAPH (C4orf26). The protein encoded by ODAPH (OMIM 614829 [106]; 4q21.1) is proline-rich, has a signal peptide and, when phosphorylated at the C-terminus, promotes hydroxyapatite nucleation and growth in vitro [6, 75]. Biallelic loss causes hypomineralised AI (AI2A4, OMIM #614832 [38]), identified in an Omani family and in eight additional families from a panel of 57 recessive families; the variants are nonsense, frameshift or splicing variants that escape NMD, and the enamel contains thin crystallites typical of early maturation [6, 75]. The variant NM_178497.5:c.229C>T p.(Arg77Ter) was found in four Omani families [6, 87].

ACP4. Testicular acid phosphatase (OMIM 606362 [107]; 19q13), encoded about 111 kb proximal to KLK4, is expressed in secretory ameloblasts and in the brain, and its function in amelogenesis is unknown [6]. Biallelic loss causes hypoplastic AI (AI1J, OMIM #617297 [33]), described in 2016, with enamel about ten times thinner than normal but well mineralised; the variants are mainly missense in the extracellular domain [6, 81]. In 2026, the Leeds series added seven families, bringing the total to 22 published families; the variant NM_033068.3:c.254T>C p.(Leu85Pro) is a founder variant in three British families of Pakistani origin with an identical haplotype, and c.713C>T p.(Ser238Leu), found in four Turkish families, is a candidate founder variant in that population [87, 108]. No involvement of other organs is known [108]. A patient compound heterozygous for two ACP4 variants also had agenesis of the upper third molars and a heterozygous WNT10A variant (NM_025216.3:c.682T>A p.(Phe228Ile)), compatible with a digenic contribution to the dental phenotype [108].

RELT. The orphan receptor of the tumour necrosis factor superfamily encoded by RELT (OMIM 611211 [109]; 11q13.4) is expressed in secretory ameloblasts, and its extracellular domain is shed by ADAM10 [5]. Biallelic loss causes hypocalcified AI (AI3C, OMIM #618386 [43]), with soft, rough, discoloured enamel, post-eruptive breakdown and rapid occlusal attrition, classically leaving a ring of intact enamel on the lateral surfaces of the molars [5, 82]. A missense variant is a founder variant in the Pakistani population of the United Kingdom [5, 110]. The short stature and severe infections described in two probands of the original report were not confirmed in later series [82, 110, 111].

SP6. The zinc finger transcription factor SP6 (epiprofin; OMIM 608613 [112]; 17q21.32) is required for proliferation of the inner enamel epithelium and its differentiation into ameloblasts [83]. In 2020 a Caucasian family was described with dominant hypoplastic AI and the variant NM_001258248.2:c.817_818delinsAA p.(Ala273Lys), at a DNA-binding residue of the first zinc finger, with reduced binding to the proximal AMBN promoter; in 2021, a second family had the de novo variant c.817_818delinsAT p.(Ala273Met), with a more severe phenotype and an almost undetectable mutant protein [83, 87, 113]. The AI1K entry (OMIM #620104 [34]) describes, in the four-generation British family, nine affected individuals with generalised hypoplastic AI with an irregular surface and, in the 9-year-old boy with the de novo variant, small, spaced teeth, rocket-shaped canine and premolar roots, enlarged pulp chambers and taurodontism of the first permanent molars [34].

2.6. Master regulators and syndromic forms of known cause

FAM20A and enamel-renal syndrome. The FAM20A gene (OMIM 611062 [114]; 17q24.2) encodes a Golgi pseudokinase without useful catalytic activity that binds FAM20C, the kinase responsible for phosphorylating most secreted proteins, including those of the enamel matrix, and activates it allosterically [5, 6]. Biallelic loss removes this activation and produces failure of enamel formation with ectopic mineralisation in the pulp, gingiva and kidney; the Fam20a-null mouse reproduces the AI and severe ectopic renal calcification [5]. FAM20A is expressed in ameloblasts only in the secretory stage, which explains the hypoplastic phenotype [6]. AI with gingival fibromatosis (2011) and enamel-renal syndrome (2012–2013) were recognised as the same disease, now designated enamel-renal syndrome (ERS; AI1G, OMIM #204690 [31]) [6, 115]. Nephrocalcinosis is reported variably, which is attributed to age, genetic modifiers and exposure to calcium channel blockers [6]. In 2026, a Lyon series showed hypophosphataemia with raised fibroblast growth factor 23 (FGF-23) and low tubular phosphate reabsorption, which suggests an FGF-23-mediated renal mechanism [116]. In the AI LOVD, FAM20A was the second most common cause overall (41 families, 15.2%) and the leading cause of recessive AI [6].

The 2025 review on FAM20C and FAM20A places enamel-renal syndrome and Raine syndrome in a single group of oro-renal disorders: FAM20A, a secreted pseudokinase, is required for optimal FAM20C activity, and the two syndromes share AI, gingival fibromatosis and ectopic mineralisation [117]. The substrates phosphorylated by FAM20C with a role in normal and ectopic mineralisation and in periodontal health include osteopontin, periostin and fetuin, and gingival fibromatosis is described as a fibrocalcifying disease directly linked to dysfunction of these proteins [117]. Activation of FAM20C by bioactive lipids such as sphingosine and modulation of the two kinases as a therapeutic target remain research hypotheses [117].

DLX3: AI with taurodontism and tricho-dento-osseous syndrome. The homeodomain transcription factor DLX3 (OMIM 600525 [118]; 17q21.33) is expressed in the placenta, skin, hair, teeth and bone; in ameloblasts it peaks at the end of the secretory stage and binds the enhancers of Amelx, Enam and Odam [6]. Heterozygous variants cause tricho-dento-osseous syndrome (TDO, OMIM #190320 [45]) and hypomaturation-hypoplastic AI with taurodontism (AI4, OMIM #104510 [44]), which form a variant-dependent spectrum of severity [6, 85, 119]. The classic 4 bp deletion NM_005220.3:c.571_574del p.(Gly191ArgfsTer66), described in 1998, causes complete TDO; the 2 bp deletion c.561_562del p.(Tyr188GlnfsTer13), initially associated with AI with taurodontism, causes an attenuated form of TDO, which indicates that the C-terminus of the protein is decisive [85, 87, 119]. The AI LOVD protein notation p.(Y166Qfs*13) for this latter variant does not match validation on the MANE transcript [6, 87]. The same 2 bp allele varies between families from AI with mild taurodontism to TDO, and a heterozygous whole-gene deletion with AI, taurodontism and attenuated TDO indicates that reduced dosage is sufficient to cause disease [5, 6, 120]. Up to 2017 six variants had been described: two frameshifts that escape NMD and alter the C-terminal transactivation domain, and four missense variants in the homeodomain (residues 133 to 182) [6]. Homeodomain variants include the two missense variants identified in two Finnish families, one of them NM_005220.3:c.398G>C p.(Arg133Pro), and the de novo variant c.533A>G p.(Gln178Arg) in a Chinese patient with typical TDO; the Finnish authors concluded that TDO results essentially from loss of function and haploinsufficiency of DLX3 [87, 121–123]. In 2019, the Leeds cohort added three alleles in British families with AI and attenuated TDO: c.574del p.(Glu192ArgfsTer66), in the last exon, which escapes nonsense-mediated decay and replaces 65 C-terminal amino acids; c.476G>T p.(Arg159Leu), in the homeodomain; and a heterozygous deletion of the entire coding region [87, 124]. In 2026, a series from Taiwan described the first splicing variant, c.516+1_516+2insA, which in a minigene assay generated two aberrant transcripts, with products p.(Val173AspfsTer28) and p.(Arg120_Val173del) according to the authors, mainly cytoplasmic in localisation and with greatly reduced transactivation [87, 125]. In the same work, DLX3 directly activated WNT10A through a conserved enhancer and three upstream binding sites, in human dental pulp cells, which the authors propose as the basis of the dentine defect and taurodontism [125]. In the mouse model with conditional deletion of Dlx3 in the dental epithelium, the enamel of all teeth was hypomineralised and chalky without significant change in matrix proteins or proteases; the ion transporters and carbonic anhydrases of the maturation stage were altered, with loss of enamel pH oscillations, and most of these genes had DLX3 binding at the proximal promoter [126]. In bone, the 4 bp variant increases mineral density in both intramembranous (calvarial) and endochondral bone: in a prospective study of 34 members of families with the same variant (20 affected, 14 unaffected), densitometry (dual-energy X-ray absorptiometry, DXA) showed higher bone mineral density in affected individuals at the distal radius and ulna, the femoral neck and the lumbar spine L2–L4 (p < 0.05) [127]. Cellular and murine studies cited in a 2012 review attribute to the 4 bp variant an increase in osteoblastic differentiation and a reduction in interferon gamma-mediated osteoclastic resorption, in contrast with the pro-apoptotic effect on odontoblasts, which explains the smaller amount of dentine and the large pulp chambers [128].

Other syndromic forms. In Jalili syndrome, CNNM4 (OMIM 607805; 2q11.2) encodes a mediator of basolateral magnesium extrusion, and it has been proposed that the retinal dystrophy and the enamel defect both result from disturbed ion homeostasis [5, 129]. In Kohlschütter-Tönz syndrome, ROGDI encodes a protein identified in 2025 as a subunit of the Rabconnectin-3 complex, involved in assembly of the vacuolar-type proton ATPase, and SLC13A5 encodes the citrate cotransporter NaCT, whose absence increases extracellular citrate and impairs mineralisation of the enamel, of which citrate is a constituent [5, 130]. In Heimler syndrome, hypomorphic PEX1 and PEX6 variants produce the mildest end of the Zellweger spectrum [5, 131]. In LTBP3, biallelic loss disturbs storage and activation of transforming growth factor beta and causes dental anomalies with short stature, whereas heterozygous variants cause geleophysic or acromicric dysplasia without agenesis [5, 132]. In CRAC channel (calcium release-activated calcium channel) channelopathies due to ORAI1 or STIM1, store-operated calcium entry is required for enamel mineralisation and the defect affects maturation [5, 133]. In non-lethal Raine syndrome, loss of FAM20C reduces phosphorylation of amelogenin, amelotin, enamelin and osteopontin and increases FGF-23 [5, 134]. In autoimmune polyendocrine syndrome type 1 (AIRE gene; OMIM #240300 [135]), enamel hypoplasia is common and independent of hypoparathyroidism, and in coeliac disease enamel hypoplasia is associated with autoantibodies against ameloblast-specific proteins, which has led to the description of an autoimmune AI in both conditions [5, 136].

Recurrent variants in enamel-renal and Kohlschütter-Tönz syndromes. In enamel-renal syndrome, eleven patients from five Brazilian families were homozygous for the same single-nucleotide deletion in exon 11 of FAM20A, NM_017565.4:c.1447del p.(Glu483LysfsTer24) (published on the genomic sequence NG_029809.1 as c.1447delG), interpreted by the authors as the first founder effect described for this gene [137]. In two Egyptian families the homozygous variants NM_017565.4:c.915_918del p.(Phe305LeufsTer76) and NM_017565.4:c.1219+3_1219+6del were identified, the latter with a protein effect that cannot be predicted without a study of RNA processing [138]. In SLC13A5-related disease, more than 40 biallelic variants have been described (nonsense, missense, frameshift, exon and whole-gene deletions), in the homozygous or compound heterozygous state; the most frequent are NM_177550.5:c.655G>A p.(Gly219Arg) and NM_177550.5:c.680C>T p.(Thr227Met), and patients with the same variants differ in seizure frequency and degree of disability, which suggests the involvement of modifiers [139]. In the North American prospective cohort, the genotype–phenotype correlation for development was limited [140]. In ROGDI, loss-of-function variants predominate, such as NM_024589.3:c.286C>T p.(Gln96Ter), c.402C>G p.(Tyr134Ter), c.469C>T p.(Arg157Ter) and the intronic deletion NM_024589.3:c.45+37_46-30del (published as c.46-37_46-30delGGCGGGGC) [141].

2.7. Relative frequency of the genes

The AI LOVD, a database of published variants maintained in Leeds, contained on 23 May 2017 192 variants in 270 families with non-syndromic AI [6]. Four genes accounted for 163 families (60.4%), and inheritance was autosomal dominant in 132 families (48.9%), recessive in 109 (40.4%) and X-linked in 31 (11.5%) [6]. The figures describe published families with an identified variant and are biased by the recruitment of consanguineous families, by studies focused on particular populations, by the time since discovery of each gene and by under-reporting of carriers of JEB variants; they are not population frequencies [6].

Source: AI LOVD, 23 May 2017 [6]; protein notations corrected and MANE equivalents by VariantValidator (GRCh38; MANE Select transcripts NM_198488.5, NM_017565.4, NM_031889.3, NM_001142.2, NM_005220.3, NM_004771.4, NM_182758.4, NM_178497.5, NM_000228.3, NM_004917.5, NM_033068.3) [87]. One family had variants in COL17A1 and ODAPH and another in COL17A1 and LAMA3, so the total is 270 and not 272 [6].

2.8. Penetrance, expressivity and allelic disorders

Expressivity varies within and between families, especially in dominant forms: in some families the phenotype is constant, in others there are substantial or subtle differences between affected individuals [1]. In females heterozygous for AMELX, the proportion of affected enamel depends on the skewing of X inactivation [5]. In FAM83H-related AI, incomplete penetrance has not been described [101]. In genes with two inheritance patterns — ENAM, AMBN, LAMB3, LAMA3, COL17A1 — the effective inheritance depends on the dosage and mechanism of the variant, which makes it necessary to interpret each result in the context of the family [6, 7, 91]. The main allelic disorders are JEB (LAMB3, LAMA3, COL17A1), epithelial recurrent erosion corneal dystrophy (heterozygous gain-of-function COL17A1 variants), dRTA (WDR72), TDO (DLX3), geleophysic and acromicric dysplasia (heterozygous LTBP3) and the Zellweger spectrum (PEX1, PEX6) [5, 14, 15].

3. Clinical Spectrum and Manifestations by System

3.1. Enamel phenotypes

The dental picture defines the class and varies with the phenotype. In hypoplastic AI, the enamel is reduced in thickness but hard and more radiodense than dentine; the surface may be smooth, rough or pitted, and the teeth appear small, yellowish (because the dentine shows through the thin enamel) and spaced, without the contours of a normal volume of enamel [6, 22, 88]. When the defect is one of quantity only, the enamel retains its translucency and does not undergo significant attrition [1]. In hypomaturation AI, the enamel is of normal thickness but opaque, white to yellow-brown or mottled, with an “agar jelly” appearance, with brown pigment in the middle layers; it chips and wears after eruption, exposing the dentine, and radiographic contrast with dentine is reduced [5, 6, 22]. Mild hypomaturation forms, with a whitish surface, may be confused with fluorosis [1]. In hypocalcified AI, the enamel is of normal thickness at eruption but soft, rough and discoloured, is rapidly lost and, in the most severe forms, has a “cheesy” consistency, leaving reduced and very sensitive teeth [1, 22]. In the AAPD reference figure, the hypocalcified form has normal thickness, a smooth surface and reduced hardness; pitted hypoplastic, normal thickness, pits and normal hardness; generalised hypoplastic, reduced thickness, smooth surface and normal hardness; and hypomaturation, normal thickness, chipped surface, reduced hardness and opaque white colour [22]. Hypoplastic and hypomineralised defects often coexist in the same patient and even in the same tooth, and the phenotype evolves: hypomineralised enamel may, after post-eruptive breakdown, appear hypoplastic [1, 2, 6]. The primary dentition may appear less affected than the permanent dentition [23].

3.2. Associated oral manifestations and functional impact

AI is accompanied by other dental and jaw changes with variable frequency. The AAPD lists accelerated or delayed eruption, low caries susceptibility, rapid attrition, excessive calculus deposition, gingival hyperplasia, enlarged follicles, impacted permanent teeth, ectopic eruption, agenesis (including of second molars), coronal or root resorption, pulp calcification and, rarely, enamel resorption and ankylosis [22]. The British consensus adds taurodontism, root dilaceration and idiopathic coronal resorption, and notes that calculus is more abundant in hypomineralised forms, especially the hypocalcified form [23]. In a British study of 30 children with AI (70% with hypomineralised enamel, 30% hypoplastic), 22 (73.3%) had at least one of six anomalies investigated (pre-eruptive coronal resorption, eruption disturbances, pulp calcifications, taurodontism, gingival enlargement, anterior open bite); dental age was advanced in 55.2% and delayed in 41.4%, but in 89.7% it differed by less than 18 months from chronological age [142]. ERN CRANIO also mentions transverse discrepancies of the maxillary arch, greater caries susceptibility in some forms, chronic gingivitis and follicular cysts, the latter mainly in enamel-renal syndrome [20].

Anterior open bite is the most studied malocclusion. The 2007 Orphanet review reports a skeletal anterior open bite in about 50% of patients with X-linked or autosomal AI; the AAPD gives 50% in hypoplastic AI, 31% in hypomaturation and 60% in hypocalcified AI; a 2023 systematic review of 339 patients found open bite in 27.7% (35.5% in hypoplastic and 36.7% in hypomineralised phenotypes), with no established genotype–phenotype correlation [1, 22, 68]. The differences reflect different samples and definitions and do not allow a single figure. The French PNDS reports malocclusions in 43% of patients with AI, without stating the source cohort [19]; the figure matches that of a Chilean series of 72 patients from 41 families, recruited between 2003 and 2016, with malocclusions in 31 (43%), anterior open bite in 12 (17%), crossbite in 7 (10%) and no significant difference between clinical subtypes (from 29% in the hypomature form to 80% in the 5 patients with the hypocalcified form) [143]. The PNDS includes in the associated oral picture microdontia, root morphology anomalies, intrapulpal calcifications, eruption disturbances, anterior open bite with mandibular hyperdivergence, posterior open bite, sagittal and transverse discrepancies and gingival hyperplasia [19].

Thermal and osmotic hypersensitivity is marked when hypoplasia predominates and in hypomineralised forms, even with enamel of full thickness, and makes toothbrushing difficult, which aggravates plaque and gingivitis [1, 20, 23]. Progressive loss of the vertical dimension of occlusion results from wear and breakdown of the enamel [20]. As enamel does not regenerate, the defect is permanent and the deterioration is mechanical, not metabolic [1, 5].

3.3. Psychosocial impact and quality of life

The psychosocial impact is considerable and was long underestimated [84]. Patients report covering their teeth with paper or chewing gum to appear “normal”, adolescents withdraw socially and there have been suicide threats because of the teeth; many young people request extraction and dentures [1]. A North American study cited by the AAPD found that adults with AI had fewer long-term relationships and fewer children than unaffected people [22, 144]. Bullying, costs, frequent appointments and emergency visits weigh on families, and access to specialist care is limited by unclear referral pathways and distance from centres [20]. In a British service, a specific 9-question questionnaire administered to 68 children and young people aged 7 to 19 years showed that those under 13 mainly reported pain and difficulty eating and adolescents mainly bullying and lack of confidence; satisfaction with appearance rose from 26% to 42% midway through treatment [145]. In 30 Māori children and adolescents with Polynesian AI, compared with 60 matched controls, oral health-related quality of life was significantly worse after adjustment for age and deprivation [21]. The French PNDS highlights the ages of greatest vulnerability: school entry, when teasing by peers can have a lasting effect on self-esteem and social relationships; adolescence, with aesthetic and functional concerns; and, in adulthood, entry into working life, life as a couple and the wish to have children [19, 144]. Parents who transmitted the disease often report guilt and shame, as well as stress associated with dental care, and lack of a diagnosis or the dismissal of the problem by oral health professionals are additional sources of distress [19].

3.4. Picture by gene in non-syndromic forms

Polynesian AI is a moderate to severe hypomaturation form described in people of Polynesian ancestry, including the Māori of Northland, New Zealand: bilaterally symmetrical hypomineralised dentition, mottled or uniformly yellow to brown, with a gradient of severity from anterior to posterior, chalky appearance, post-eruptive breakdown and hypersensitivity; expression in the primary dentition is poorly known and prevalence in the Māori and Polynesian population is unknown [21]. The 2007 Orphanet review mentions a high frequency of a recessive allele in some Polynesian communities [1].

3.5. Enamel-renal syndrome (FAM20A)

The oral profile is distinctive and, according to the authors who defined it in 2014, pathognomonic: generalised hypoplastic AI with thin or absent enamel in both dentitions (sometimes with “glassy”-looking incisors), delayed or failed eruption with multiple impacted teeth, intrapulpal calcifications, hyperplastic dental follicles visible as pericoronal radiolucencies, root anomalies, hypodontia in some patients and gingival hyperplasia that may contain calcified nodules [2, 5, 6, 115]. The renal component is bilateral nephrocalcinosis and/or nephrolithiasis, asymptomatic in childhood and detected only when looked for, with plasma calcium usually normal [5, 146]. A 2021 systematic review gathered 69 patients aged 1 to 64 years, with no sex predominance, consanguinity in about a quarter and a FAM20A variant in all those tested [5, 146]. In Lyon, four homozygous children (two consanguineous families) referred by the dental team at 14.5 years (11–19) had no haematuria or colic, but all had multiple bilateral nephrolithiasis and one family also had nephrocalcinosis; renal function was preserved (estimated glomerular filtration rate 109 mL/min/1.73 m², range 93–111), with hypophosphataemia (−1.9 standard deviations for age), raised FGF-23 (98 RU/mL; normal 21–91), hypocalciuria, low maximal tubular reabsorption of phosphate (TmP/GFR), parathyroid hormone at the upper limit (83 ng/L; local limit 65) and 25-hydroxyvitamin D of 44 nmol/L (19–92) [116]. Two adult siblings with a homozygous deletion of exon 4 of FAM20A had recurrent calcium nephrolithiasis without nephrocalcinosis on CT, hypocalciuria and marked hypocitraturia, with normal serum calcium and phosphate, which extends the renal phenotype to stone-forming forms without nephrocalcinosis [147]. The disease described as AI with gingival fibromatosis corresponds to the same entity in patients without renal assessment [5, 6].

The largest series from a single population gathered 11 patients from five Brazilian families, aged 6 to 25 years, all homozygous for the same FAM20A variant: all had hypoplastic AI, microdontia, intrapulpal calcifications, impacted posterior teeth with hyperplastic pericoronal follicles, gingival fibromatosis, ectopic gingival and pericoronal calcifications and nephrocalcinosis; only four had laboratory abnormalities (vitamin D, parathyroid hormone, phosphate or calcium), and intellectual disability and renal cysts occurred in two patients each [137]. In two Egyptian families, the main complaints were AI and gingival hyperplasia; gingival biopsy showed psammomatous calcifications, ultrasound revealed bilateral nephrocalcinosis in both patients, and craniofacial features not previously reported were described (long face, thick lips, notched upper central incisors and thick alveolar ridge) [138]. A 2025 scoping review, with 62 studies published between 1972 and 2024, confirmed the constancy of the oral profile and found that renal, craniofacial and systemic manifestations are reported heterogeneously, with no long-term follow-up studies defining long-term renal risk; until such data exist, the authors advocate monitoring of renal function [148]. Nephrocalcinosis is frequent in the series in which it is looked for, but its course in adulthood and the risk of chronic kidney disease are not quantified in the sources consulted [116, 148].

3.6. Hypomaturation AI with distal renal tubular acidosis (WDR72)

A subset of patients with WDR72-related AI has hereditary dRTA, which usually presents in childhood with poor weight gain and growth failure, vomiting, polyuria, polydipsia and constipation, hyperchloraemic metabolic acidosis with a normal anion gap, hypokalaemia (potassium <3.5 mEq/L), inappropriately alkaline urine, hypercalciuria and hypocitraturia, with nephrocalcinosis, nephrolithiasis, medullary cysts, impaired renal function and bone demineralisation [5, 14]. Sensorineural deafness belongs to the forms due to ATP6V0A4, ATP6V1B1 and FOXI1, and haemolytic anaemia to the SLC4A1 form; neither is part of WDR72-related disease [5]. In the Lyon series, two of the 30 patients with AI screened by renal ultrasound had WDR72 variants [116]. The combination of dRTA with abnormal enamel points to WDR72 [5]. An isolated case of dRTA due to SLC4A4 with autoimmune thyroiditis, tooth agenesis, enamel hypomaturation and pulp stones suggests that other renal acidification genes may be associated with enamel defects [14].

3.7. Tricho-dento-osseous syndrome and AI with taurodontism (DLX3)

TDO was defined by Lichtenstein and colleagues in 1972 in 107 individuals from a single family, after Robinson had described in 1966 a dominant disorder with curly hair and enamel hypoplasia [120]. It combines kinky hair from birth, which sometimes straightens with age; small, spaced, pitted teeth with poor enamel and taurodontism in both dentitions, markedly delayed or advanced dental maturity and pulps filled with amorphous material; frontal bossing, dolichocephaly, a square mandible with a retruded maxilla and relative prognathism; mild to moderate increase in bone density and thickness, especially of the calvarium, with no mastoid pneumatisation, obtuse mandibular angles and a short ramus; brittle nails with peeling in about half; and delayed bone age [120, 149]. The teeth erode and discolour, develop periapical abscesses and are lost by the second or third decade; intellect is normal and there is no increase in fractures [120, 149]. In a 2025 systematic review with 27 studies and 297 individuals, the most frequent findings were taurodontism (70.5%), increased bone density or thickness (43.8%), enamel hypoplasia (34.5%), dental infections (28.7%) and sparse hair (27.2%); the pooled meta-analytic estimates were 73% for taurodontism and 71% for enamel hypoplasia (95% confidence interval, 95% CI, 0.52–0.97 for both) [150]. The AI4 form has similar dental findings with normal or minimally affected hair and bones [85, 149].

The findings are distributed differently with age. Kinky hair is present at birth in about 80% of patients, and about half of these lose the hair phenotype by adolescence; the dental changes are manifest from the primary dentition, whereas the bone changes become more marked in adulthood, with increasing variability of mineral density in the radius and ulna and increasing vertebral density with age [18, 128]. In a North American series cited in the 2012 review, the bone changes were of variable expression: cortical thickening in 65%, obliteration of the diploë in 68%, absence of pneumatisation of the frontal sinus in 49% and of the mastoid in 81% [128]. In three affected women from three generations of one family, bone density was increased in the long bones and the skull base, the mandible had dense trabeculation and there was no resorption of the alveolar ridge after tooth loss [151]. Bone thickening is not associated with fractures or other evident pathology [128, 149]. In the cephalometric study of 53 affected individuals and 34 unaffected relatives, affected individuals had a class III skeletal pattern, with smaller SNA and ANB angles due to maxillary retrusion, a longer mandibular body and a shorter ramus, but variability was great in both groups, ranging from severe class III to moderate class II with mandibular retrusion [128, 152]. Other manifestations described are dolichocephaly due to craniosynostosis, flat or brittle nails, sometimes only in some toenails, impacted teeth, delayed eruption, clinodactyly and skin lesions; in an isolated report, uncombable hair coexisted with atopic dermatitis [18, 128].

The enamel in TDO is one quarter to one eighth of normal thickness, hypomineralised and, on radiographs, less opaque than dentine, with a moth-eaten appearance; the pulp horns reach the amelodentinal junction, which explains the abscesses due to pulp microexposure after enamel attrition or fracture, present in more than half of the patients in the original family [128]. Taurodontism occurs in about 3% to 8% of the general population, mostly mild and in the upper second molar; in TDO it is severe, involves all molars and in particular the lower first permanent molar, and is, together with enamel hypoplasia, a finding with complete penetrance in carriers of the 4 bp variant [128]. In attenuated forms the picture is reduced: in a six-generation Korean family with the c.561_562del variant, the proband had no hair changes, the mother had only pulp elongation of the third molars and a cousin had curly hair at birth, the mildest phenotype described for the gene [153]. In the three British families of 2019, the picture was hypoplastic AI with taurodontism and kinky hair in childhood that normalised in adolescence, with no evident bone changes on dental radiographs; the patient without a family history also had delayed eruption of the lateral incisors with talon cusps, internal resorption of one tooth and epileptic seizures, of unestablished causal relationship [124]. In the Taiwan series, the carrier of the splicing variant had a predominantly mesenchymal phenotype, with severe dentine hypoplasia, large pulp chambers and hypertaurodontism, but near-normal enamel, whereas the mother and sister had a predominantly epithelial phenotype, with enamel hypoplasia and kinky hair [125]. In the Finnish families with homeodomain variants, dental maturation was markedly delayed or markedly advanced, the pulps were filled with amorphous denticle-like material and, in one of the families, the affected individuals developed only lanugo-type hair and severe dental changes [121].

3.8. Jalili syndrome (CNNM4)

Jalili syndrome combines cone-rod dystrophy with AI of both dentitions [129, 154]. In a series of seven patients from six families, the mean age at presentation was 6.7 years (3–16), mean initial visual acuity was around 20/246, all had nystagmus (which increases under photopic conditions) and six had photophobia; the fundus varied between mild disc pallor, vascular attenuation and macular atrophy, and could appear almost normal; electroretinography showed cone-rod dysfunction and structural and functional progression was documented in all [5, 154]. In the three original siblings from Gaza, AI was accompanied by anterior open bite [5]. In the French cohort, the AI led to the retinal diagnosis: in a 4-year-old boy followed for enamel problems, the genetic diagnosis of biallelic CNNM4 variants led to confirmation of the retinal dystrophy [2].

3.9. Kohlschütter-Tönz syndrome (ROGDI and SLC13A5)

The syndrome, described in a Swiss family in 1974 and registered by Orphanet as amelocerebrohypohidrotic syndrome, combines early-onset and often refractory epilepsy, severe global delay and AI of both dentitions, with yellow or brown teeth [5, 61]. In two independent reviews of 43 patients with ROGDI variants, epilepsy, global delay and AI were present in 100%, language disorder in all, attention deficit hyperactivity disorder in 17.2%, suspected autism spectrum disorder in 3.4% and emotional disturbances with self-injury in 9.1%; magnetic resonance imaging and electroencephalography may be normal [5]. In the SLC13A5 form (developmental epileptic encephalopathy 25 with AI), seizures begin in the first days of life (22 of 23 patients in one series), there is axial hypotonia, peripheral hypertonia and dystonia, and almost all have hypoplastic AI with yellow-orange discoloration, thin enamel that chips, wide interdental spaces, abnormal crowns, delayed eruption and hypodontia [5, 130]. Dental histology distinguishes the two causes: hypocalcified AI in the ROGDI form, hypoplastic in the SLC13A5 form [5, 130].

In the SLC13A5 form, the North American prospective natural history study (30 children and adults, frequent visits over 2 years) showed that seizure burden and emergency department visits are greatest in the first decade and decrease after 10 years, but older patients remained on an average of three antiepileptic drugs and some had intercurrent seizures; most electroencephalograms were abnormal, although fewer than a third had frequent or abundant interictal epileptiform activity, and epilepsy-related quality-of-life indices were low and stable, with greater cognitive and executive impairment than impairment of mood and behaviour [155]. In the same cohort, standardised developmental scales (Mullen, Peabody and Vineland) showed global impairment, with modest gains in early childhood and static skills in adolescence and adulthood [140]; most adults do not live independently [139]. The NaCT cotransporter is expressed in the brain, teeth, liver and testes; the dental changes (hypoplasia, hypodontia, yellow to orange coloration, abnormal crowns, wide interdental spaces and gingival hyperplasia) affect both dentitions and may require a soft diet, and in a study of 15 patients height, weight and head circumference did not differ from peers [139]. In the ROGDI form, a 6-year-old Latvian girl had idiopathic bilateral nephrocalcinosis, a finding not described in the 43 patients published until then, and super-refractory status epilepticus [156]. In a Chinese girl homozygous for the intronic ROGDI deletion, motor delay (developmental quotient of 38.8 at 7 months) preceded the first seizure, at 10 months; magnetic resonance imaging was normal, the uniform yellow coloration and caries of the teeth were noted from 1 year, and at 7 years profound global delay, language disorder and dependence in daily activities persisted [141].

3.10. Other syndromic forms

In Heimler syndrome, the mildest of the Zellweger spectrum, there is sensorineural deafness, AI with enamel hypoplasia restricted to the permanent dentition, nail changes and late or occasional retinal dystrophy, in an individual who may be neurologically normal; almost all patients with intermediate or mild forms of the spectrum have AI of the permanent dentition [2, 5, 131]. In dental anomalies with short stature (LTBP3, recessive), also designated brachyolmia with AI, the AI is hypoplastic with markedly reduced or absent enamel, with oligodontia, taurodontism, single-rooted molars, abnormal dentine and failure of mandibular eruption, and is associated with significant short stature with platyspondyly and posterior scalloping of the vertebrae; three asymptomatic homozygous siblings had mitral prolapse and aortic root dilatation detected by screening [5, 132, 157]. In junctional epidermolysis bullosa, pitted enamel dysplasia is common in all forms and persists in those compatible with adult life [15]. In non-lethal Raine syndrome (FAM20C), survivors to adulthood may have craniosynostosis, hypophosphataemic rickets and multiple dental anomalies with AI [5, 134]. In the syndrome of short stature, AI and skeletal dysplasia with scoliosis (SLC10A7), severe disproportionate short stature begins in the prenatal period and is associated with micro-retrognathia, multiple congenital dislocations and advanced carpal ossification [158]. SLC10A7 belongs to the multiple-pathway congenital disorders of glycosylation, and AI is the clinical feature that the textbook of metabolic diseases associates with this gene [159]. In the loss-of-function ORAI1 and STIM1 immunodeficiencies, severe-type combined immunodeficiency coexists with non-progressive congenital myopathy, anhidrosis and a defect of enamel maturation in both dentitions [5, 133]. In the 2018 description of the SLC10A7 syndrome, six patients of Turkish, Dutch and Iranian origin had hypomineralised AI with yellow-brown enamel with a rough surface and short, spaced crowns, and other patients had hearing loss and mild intellectual disability [52]. In the syndrome of aplasia cutis with enamel dysplasia (FOSL2, dominant), ten patients from nine families had aplasia of the scalp, sometimes with an underlying bone defect, and enamel hypoplasia with discoloration and visible dentine, with mostly de novo variants and suspected germline mosaicism in one family [57, 65]. In the syndrome of hypotonia, ataxia, developmental delay and tooth enamel defect (CTBP1, de novo variant NM_001012614.2:c.991C>T p.(Arg331Trp) in four patients), all had soft, discoloured enamel [58, 66]. Other syndromes with AI or prominent enamel hypoplasia include Shaheen syndrome (COG6, with hypohidrosis and palmoplantar keratoderma) and USP7-related disease, in which about 55% of patients have dental problems, including AI [4, 160].

3.11. Natural history and treated disease

Data on the untreated natural history are mainly descriptive: affected enamel deteriorates mechanically with use, the teeth become sensitive, lose structure and vertical dimension, and in severe forms, such as TDO, are lost by the second or third decade [1, 120]. Data on treated disease come mainly from restorative studies in children and young people: after any type of restoration, pain and hypersensitivity decrease and aesthetics improve, and ceramic crowns placed in 27 adolescents and young adults had a survival of 97.8% after 13.4 years, with sensitivity reduced from 4.6 to 0.9 on a visual analogue scale [84, 161]. The two types of data are not directly comparable, and the course of extradental manifestations follows that of the underlying syndrome.

4. Clinical Recognition and Red Flags

4.1. When to suspect

Suspicion arises from the appearance of the teeth and their distribution. Four simple questions, proposed in the Orphanet review, guide the reasoning: are there other relatives with similar changes; are all teeth affected in a similar way; is there a chronological distribution of the lesions; and was there any illness in the medical history capable of disturbing enamel formation [1]. A generalised, symmetrical defect affecting both dentitions with a family history suggests AI; a defect that follows the chronology of tooth formation, spares groups of teeth formed at another time or is asymmetrical suggests an environmental cause or MIH [1, 22, 68]. In a survey of 311 dentists, 38.7% of general practitioners reported difficulty in distinguishing AI from MIH, a difficulty less frequent among paediatric dentists [68].

By age, AI can be recognised in the primary dentition from eruption of the first tooth, around 6 months, and ERN CRANIO recommends that the first examination take place at that time [20]. Eruption of the first permanent teeth, from 6 years, is a second moment of recognition, especially in forms that affect the permanent dentition more, such as Heimler syndrome and Polynesian AI [1, 2, 21]. Delayed eruption with impacted teeth, prolonged retention of primary teeth and anterior open bite are reasons for referral in adolescence, as in the two young Brazilian women with FAM83H-related AI referred at 15 and 18 years [101].

The French PNDS describes three usual circumstances of suspicion: observation of enamel defects in the primary or permanent dentition, often prompted by parental concern about discoloration from eruption of the first teeth; identification of AI in a syndromic picture already known or under investigation, with referral to confirm enamel involvement; and a family history of AI, which leads families to seek consultation early to confirm or exclude the defect in the child [19].

4.2. Signs pointing to a syndromic form

AI may be the first sign of a systemic disease, and the dental phenotype sometimes acts as a diagnostic signature [2, 6]. The following signs justify targeted assessment.

Examination of the teeth is part of the assessment of children with epilepsy and developmental delay, as it may reveal enamel hypoplasia of syndromic cause [162]. The 2007 review already acknowledged the need for renal examination whenever a diagnosis of AI is made, because of the association with nephrocalcinosis [1]; the Lyon centre began in 2022 to refer all patients with AI for renal ultrasound, and among 30 patients screened found renal involvement in six (four with FAM20A and two with WDR72) [116].

4.3. Emergency signs and newborn screening

AI does not cause medical emergencies of its own; dental emergencies arise from pain, hypersensitivity, fractures and abscesses, more frequent with taurodontism and in TDO [20, 120]. In syndromic forms, emergencies belong to the organ involved (for example, severe hypokalaemia in dRTA or status epilepticus in Kohlschütter-Tönz syndrome) [5, 14]. There is no newborn screening for AI; screening for X-linked adrenoleukodystrophy by C26:0-lysophosphatidylcholine may detect the Zellweger spectrum but misses mild forms such as Heimler syndrome [5].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Clinical criteria

There are no formally validated criteria. Clinical diagnosis requires excluding extrinsic, environmental and chronological factors, establishing a probable inheritance pattern, recognising the phenotype and comparing it with the dates of tooth formation [1]. The family history, pedigree, meticulous clinical observation and photographic records form the basis, and the British consensus adds, before any plan, radiographs (quantity of enamel, associated anomalies, baseline for follow-up), pedigree, genetic testing when available, photographs and a patient-reported outcome questionnaire [1, 23]. ERN CRANIO recommends that the first appointment gather the medical and dental history, complete clinical and radiological documentation, quality-of-life assessment, digital scanning, three-dimensional photographs and DNA testing, and that the diagnosis be confirmed by a rare disease specialist and/or geneticist from 2 years of age, with genetic testing especially when a syndrome is suspected [20]. The French PNDS bases the diagnosis on the family history, pedigree and meticulous clinical and radiographic examination, carried out in liaison with a reference or competence centre of the O-Rares network; the variability of expression justifies assessment by a dentist experienced in these disorders [19].

For TDO, the only proposed criteria found in the sources are those of Seow, taken up in a 2012 review: generalised enamel defects, of hypomaturation or hypocalcification associated with hypoplasia, confirmed clinically and radiographically; severe taurodontism involving the lower first permanent molars; autosomal dominant inheritance; and at least one minor criterion (brittle or peeling nails, bone sclerosis on the lateral cephalometric radiograph, or kinky, curly or wavy hair in childhood, verifiable in photographs, which may straighten later) [128]. The degree of taurodontism of the lower first molar can be measured on the panoramic radiograph as the ratio between crown–body height and root length, measured along the axis of the tooth from the occlusal fossa to the furcation and from the furcation to the apices, which allows the teeth to be classified into four types [128]. These criteria have not been validated against genetic testing and do not cover attenuated forms without hair, nail or bone involvement, which are recognised by identification of a DLX3 variant [123, 124].

5.2. Radiographic assessment

Intraoral radiographs show the contrast between enamel and dentine and the degree of hypoplasia; extraoral radiographs may reveal impacted teeth and spontaneous resorption [1]. In hypoplastic AI, the radiographic thickness of the enamel is reduced but its radiodensity remains greater than that of dentine; in hypomaturation AI, thickness is normal and radiodensity approaches that of dentine; in hypocalcified AI, radiodensity is equal to or lower than that of dentine [5]. The panoramic radiograph is essential in the primary dentition to identify anomalies of eruption and number [20], and cone-beam computed tomography shows enlarged follicles and pre-eruptive coronal resorptions that conventional radiography does not show [101]. The French PNDS tailors the initial imaging to age and anomalies: panoramic radiograph for dental anomalies; lateral cephalometric radiograph with cephalometric analysis for orthodontic assessment and posteroanterior cephalometric radiograph when there is a transverse anomaly or facial asymmetry; and cone-beam tomography, preferably with a low-dose protocol, for complex anomalies, doubts in interpreting the panoramic radiograph, delayed eruption, impactions, marked asymmetry or when orthodontic-surgical treatment is anticipated; before ordering investigations, previous radiographs and their dates are sought [19]. The signs to look for are reduced, absent or irregular enamel thickness, abnormal radiopacity, shape anomalies (taurodontism, roots), dentine anomalies and intrapulpal calcifications, and changes in eruption, the periodontium and the alveolar and basal bone [19]. In surveillance of the rehabilitated adult, the PNDS prefers interproximal radiographs (bite-wing) to the panoramic radiograph, as they detect incipient carious lesions better [19]. Scanning electron microscopy, microtomography and microhardness of exfoliated teeth are research tools and not clinical tests [5].

5.3. Genetic diagnosis

The German S3 guideline recommends genetic diagnosis, usually with a genetics consultation, when the clinical and radiological findings suggest a genetic cause (recommendation 2, expert consensus, moderate strength) [24]. The 2024 Swedish review concludes that recent genetic evidence makes it advisable to recommend testing to complement clinical diagnosis [84]. The phenotype does not allow the gene to be inferred, because more than 70 genes produce overlapping pictures and the same gene may produce different phenotypes [2, 5].

The yield depends on the technique and the cohort. In a North American cohort of 71 families studied by sequencing of six candidate genes, a molecular diagnosis was obtained in 26 families (37%) and 132 affected individuals (59%): FAM83H 12 families (46%), AMELX 6 (23%), ENAM 3 (11%), KLK4 2 and MMP20 2 (8% each) and WDR72 1 (4%) [102]. Sanger studies of candidate genes found variants in 36.6% to 48.7% of families, and a focused clinical exome diagnosed 18 of 65 cases of syndromic and non-syndromic AI (27.7%) [6]. In the French O-Rares cohort (2009–2021), the GenoDENT panel (version 6.0, 567 genes, of which 248 diagnostic and 319 candidate, capturing the exons and 25 intronic bases on each side) had a diagnostic rate of 60%; among 115 index cases from 111 families, 73% had non-syndromic AI and 27% syndromic; the diagnosis was validated with class 4 or 5 variants in 81% of the cohort and variants of uncertain significance remained in 19% of index cases; the most frequent genes were MMP20 and FAM83H in isolated AI and FAM20A and LTBP3 in syndromic AI, and negative cases were resolved by exome (for example, ACP4 or digenic inheritance) [2]. In the United Kingdom, the National Health Service (NHS) has a diagnostic AI panel [6, 23].

In France, the PNDS still considers Sanger sequencing a reference method, but records the move to dedicated next-generation sequencing panels, such as GenoDENT, and accepts exome and genome sequencing; analysis is generally performed from the outset as a trio (child and parents), in accredited medical genetics laboratories, from blood or saliva, and should preferably be prescribed by a geneticist or an organ specialist of the rare disease network, able to explain the implications to the patient and family [19]. When the panel is inconclusive and the indication is justified, the family may access, in consultation with the O-Rares network and the local genetics clinic, genome sequencing under the France Genomic Medicine Plan 2025, in the indication “syndromic forms of rare diseases with orodental expression” [19]. Suspicion of a syndromic form, from the history or examination, leads to referral to the O-Rares centres, which organise regional clinical and molecular diagnostic pathways with a geneticist [19]. The PNDS list of main genes separates those for non-syndromic AI (ENAM, AMELX, AMBN, AMTN, LAMA3, LAMB3, LAMC2, COL17A1, ITGB4, ITGB6, ACP4, RELT, ODAPH, SLC24A4, MMP20, KLK4, FAM83H, DLX3, GPR68) from those for syndromic AI (WDR72, LTBP3, CNNM4, GALNS, ROGDI, SLC13A5, SLC10A7, FAM20A, FAM20C, CLDN16, CLDN19, PEX1, PEX6, DLX3, STIM1, ORAI1, PORCN); it also includes LAMC3 and COL7A1 among the non-syndromic genes [19], which does not match the phenotype of these genes: biallelic LAMC3 variants cause occipital malformations of cortical development, with no enamel defect described [163], and COL7A1 causes dystrophic epidermolysis bullosa, in which the dental lesions are mainly secondary (caries favoured by microstomia and a soft diet), with enamel of normal chemical composition, and the occasional hypoplasia of the recessive form occurs within a syndromic skin picture [164]. The basement membrane-related forms of AI correspond to LAMA3, LAMB3, LAMC2 and COL17A1, already included in the list.

5.4. Limitations of the tests

A negative result does not exclude AI of genetic cause. Sanger and exome studies have underestimated intronic, regulatory and structural variants, and heterozygous copy number variations larger than one exon are not detected by Sanger or by exome without specific analysis; deletions of AMELX, AMBN, AMTN and FAM20A have been described [5, 6, 147]. RNA studies, useful in other diseases, are hampered by the expression of many genes being restricted to the enamel organ, which is inaccessible in the patient [6]. Interpretation requires attention to nomenclature: the AI LOVD and older articles use outdated transcripts and contain incorrect protein notations (for example, FAM83H c.973C>T, ENAM c.1259_1260insAG, DLX3 c.561_562del), and in AMELX the MANE transcript (NM_001142.2) differs in numbering from the historical transcript NM_182680.1 [6, 87]. Variants of uncertain significance should not be used for clinical decisions [2]. In females with X-linked AI, the phenotype may be mild and the proportion of affected enamel depends on X inactivation [5]. In Heimler syndrome, peroxisomal biochemistry (very-long-chain fatty acids) is often normal and does not exclude the diagnosis [5]. In distinguishing the ROGDI and SLC13A5 forms, markedly raised plasma and urinary citrate supports SLC13A5: in one infant, plasma citrate was 820 µmol/L (reference 19–83) and urinary citrate 5615 mmol/mol creatinine (reference 162–2200), which allowed a variant of uncertain significance to be reclassified as likely pathogenic [5].

The report should state the gene, the transcript with version (preferably MANE Select), the HGVS (Human Genome Variation Society) nomenclature in complementary DNA and protein, zygosity, the class according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) and, for class 4 or 5 variants, the family segregation study, as the Strasbourg laboratory does [2]. In genes with two inheritance patterns, the report should state the position of the variant (for example, last exon of LAMB3) and the expected mode of inheritance [5, 6].

5.5. Systemic assessment after diagnosis

After genetic diagnosis, reverse phenotyping may reveal additional features and turn an “isolated” AI into a systemic disease [2]. Patients with FAM20A variants should be referred for specialist renal assessment and follow-up [5, 6]; in patients with WDR72, acid–base and renal assessment is necessary [5]; in patients with LTBP3, echocardiography even in the absence of symptoms [5, 157]; in patients with CNNM4, ophthalmological assessment with electroretinography [154].

The French PNDS describes the initial dental examination, to be performed as early as possible: loss of tooth substance and nature of the defect, caries, calculus and gingival or periodontal involvement, sensitivity or spontaneous pain, on chewing or on brushing, loss of vertical dimension, dysmorphosis and occlusal disturbances, aesthetic complaint and any oral dysfunction, with models (conventional or optical impressions) and intraoral photographs to document the initial state [19]. Orthodontic assessment is systematic, age-adapted and three-dimensional, with functional examination directed at mouth breathing, anterior or lateral tongue dysfunction and speech disorders, which contribute to maintaining malocclusions [19]. The search for associated signs guides the medical consultations: loss of visual acuity, altered colour vision or photophobia suggest Jalili syndrome; epilepsy with regression, Kohlschütter-Tönz syndrome; short stature, Verloes-Bourguignon syndrome; suspected nephrocalcinosis justifies renal ultrasound; and the investigations are reassessed in the light of the molecular diagnosis [19].

5.6. Algorithm

The French PNDS decision tree for molecular diagnosis, adapted from the 2019 terminology of de La Dure-Molla and colleagues, starts from enamel anomalies of both dentitions and groups the genes into isolated AI and rare diseases by system (skin, eye, bone, endocrine system, kidney, cognitive impairment, ear, nose and throat, and immunity), which helps to choose the panel and interpret variants in the light of extraoral signs [19].

5.7. Prenatal and preimplantation diagnosis

The sources consulted do not describe the use of prenatal diagnosis or preimplantation genetic testing in isolated AI, nor studies on demand for them. Identification of the familial variants is the prerequisite for any reproductive discussion, which takes place in a genetics clinic and is covered in section 8.

6. Differential Diagnosis

6.1. Acquired enamel defects

The most frequent differential is dental fluorosis, which ranges from small white spots to dense white coloration with disfiguring stains and hypoplasia [1]. It is distinguished by its chronological distribution (horizontal white bands corresponding to periods of higher intake; premolars and second permanent molars often spared), by the lack of uniformity between teeth and by a history of excessive fluoride intake, such as the habit of eating toothpaste or the local water supply [1, 22]. Chronological hypoplasia, due to systemic illness during tooth formation (prolonged diarrhoea, coeliac disease that may only be diagnosed later, antileukaemic therapy), affects only the enamel formed in that period [1, 6]. Tetracycline staining is another non-genetic cause to exclude in apparently sporadic cases [1]. A localised defect of a central incisor accompanied by non-eruption of the adjacent tooth suggests trauma in childhood, and regional odontodysplasia affects the enamel, dentine and pulp of one segment of the arch [1].

Molar-incisor hypomineralisation is the most difficult differential for the hypocalcified and hypomaturation forms. The most useful criteria are distribution and symmetry: AI is generalised, hereditary and affects both dentitions; MIH is localised to the first molars and some incisors, asymmetrical, and is associated with hypomineralised primary second molars [1, 68]. Taurodontism is significantly more frequent in AI [68]. The British consensus notes that, unlike MIH, in AI the unerupted teeth are also affected, which changes the decision on interceptive extractions [23]. According to the French PNDS, the distinction is difficult only in severe MIH, in which the lesions may also affect the cusp tips of the canines, the second premolars and the second permanent molars; even then, MIH lesions tend to be asymmetrical, of variable severity and not involving the entire tooth surface, and hypomineralisation of the primary second molars (hypomineralised second primary molars, HSPM) is the equivalent of MIH in the primary dentition [19]. Early childhood caries is another frequent differential: rapid destruction of all primary teeth accompanies continuous and repeated intake (night-time bottle, on-demand breastfeeding) and inadequate hygiene, generally spares the lower incisors and is suggested by the feeding history; fluorosis presupposes proven chronic exposure to fluoride, especially in the first years of life [19].

6.2. Dentine disorders and other colour changes

In dentinogenesis imperfecta, the teeth are opalescent, blue-grey or brown, and normal enamel chips off an abnormal dentine; in AI, the primary defect is in the enamel and radiographs show normal or taurodont pulp chambers, not obliterated [22, 165]. Other dental discolorations of systemic cause include the green discoloration of hyperbilirubinaemia and the erythrodontia of congenital erythropoietic porphyria [165, 166]. Dentine disorders, tooth agenesis and primary failure of eruption enter the differential diagnosis.

6.3. Differential table

6.4. Differential within the class

Within AI, the phenotype guides but does not predict the gene [2]. The inheritance pattern is the most useful discriminating criterion: vertical bands in a mother with affected children suggest AMELX; vertical transmission with hypocalcified AI suggests FAM83H; pitted hypoplasia in children of patients with JEB suggests carriers of LAMB3, LAMA3 or COL17A1; consanguinity with pigmented hypomaturation suggests MMP20, KLK4, WDR72, ODAPH, SLC24A4 or GPR68 [1, 6, 15]. Associated taurodontism suggests DLX3, but it has been described with ITGB6, LTBP3 and in AI of other causes [5, 68]. In Kohlschütter-Tönz syndrome, histology (hypocalcified versus hypoplastic) and citrate distinguish ROGDI from SLC13A5 [5, 130]. In families with a mixed phenotype, ITGB6 should be considered even when the enamel does not fit a single type [5].

Gingival fibromatosis with ectopic calcifications and AI brings enamel-renal syndrome close to non-lethal Raine syndrome due to FAM20C, which is distinguished by hypophosphataemic rickets or osteosclerosis [5, 117]. In Kohlschütter-Tönz syndrome, the nephrocalcinosis described in one ROGDI case should not lead to a diagnosis of enamel-renal syndrome when there is epilepsy and global delay [156].

The boundary between TDO and AI4 has been debated. Seow argued that type IV AI should not constitute a separate entity, because the prevalence and severity of taurodontism in AI without DLX3 did not differ from those of healthy controls, and that severe taurodontism of the lower first molar is the marker of TDO [128]. In 1999, the absence of the 4 bp deletion in families with type IV AI led to the two conditions being considered genetically distinct, but the later identification of DLX3 variants in families classified as AI with taurodontism, the reclassification of the 2 bp variant as a cause of attenuated TDO and families with AI and subtle extradental signs made the two designations extremes of a single allelic spectrum [85, 124, 128]. In practice, examination of the hair in childhood photographs, of the nails and, when there is an independent clinical indication, of a lateral cephalometric radiograph is what separates an attenuated TDO from an apparently isolated AI [124, 128]. In adults, post-eruptive changes make clinical diagnosis difficult: in a French family, the mother's previous diagnosis was dentinogenesis imperfecta, and the radiopaque mandibular masses had been interpreted as florid cemento-osseous dysplasia before the DLX3 variant was identified [123].

7. Treatment and Risk Management Strategies

7.1. Principles and guidance documents

There is no treatment that corrects the enamel defect; management is dental, preventive and restorative, begins early and continues into adult life, and in syndromic forms treatment of the organ involved is added [1, 20, 24]. The aims are to reduce pain and hypersensitivity, preserve tooth structure and the vertical dimension of occlusion, restore function and normalise aesthetics, with attention to the psychological impact [20, 22]. The evidence base is weak: the German guideline rated the evidence as low or very low for most recommendations, because case reports and case series predominate, although AI is the only one of the rare dental diseases with clinical studies [6, 24].

The general recommendations of the German guideline apply to all rare dental diseases: interdisciplinary diagnosis and care throughout life, started early (recommendation 1, consensus, strong); genetic diagnosis when the clinical picture suggests it (recommendation 2, consensus, moderate); intensive, risk-oriented preventive care throughout life (recommendation 3, consensus, strong); and treatment under general anaesthesia when needs are extensive and treatment is not otherwise possible (recommendation 4, very low evidence, open recommendation) [24].

7.2. Prevention, hygiene and hypersensitivity

Prevention comes before intervention, but in AI intervention is earlier and more extensive than in other children [1]. The AAPD recommends early identification and intervention, regular periodic examinations to treat teeth as they erupt, meticulous oral hygiene, calculus removal, mouthwashes, fluoride applications and desensitising agents [22]. ERN CRANIO recommends caries risk assessment, family education on hygiene and diet, and application of high-concentration fluoride varnish even without restorative need, for prevention, remineralisation and desensitisation [20]. In the United Kingdom, children with AI are considered at high caries risk and included in an enhanced preventive regimen [23]. The French PNDS calls for an early first appointment even without restorative need, to familiarise the child with the dentist and with care, a caries risk assessment that adapts the preventive measures, brushing twice a day, supervised by an adult, with 1000 ppm fluoride toothpaste, a non-cariogenic diet and, where restorations are not needed, high-concentration fluoride varnish (Duraphat®) for prevention and remineralisation [19]. Toothpastes for sensitive teeth make brushing easier, and during treatment with fixed appliances a fluoride mouthwash is advised [19].

5% sodium fluoride varnish contains 2.26% fluoride (22,600 ppm), and acidulated phosphate fluoride gel 1.23% (12,300 ppm); meta-analyses of 23 trials, most with six-monthly application, favour varnish in both dentitions; unit doses of 5% varnish are the only professional fluoride recommended below 6 years of age for safety reasons; and children at risk of caries should receive professional treatment at least every 6 months [176]. The AAPD indicates toothpastes with at least 1000 and less than 1500 ppm fluoride and mentions prescription toothpastes with 5000 ppm [176]. These indications derive from caries prevention in general and are applied to AI by extrapolation, as ERN CRANIO and the British consensus do [20, 23].

The PNDS does not state the amount of varnish. The summary of product characteristics (SmPC) of Duraphat® (sodium fluoride 50 mg/mL, equivalent to 22.6 mg/mL fluoride, dental suspension) defines the dose per single application, performed only by a professional, as a thin layer on the most vulnerable areas, after removing plaque and drying the teeth [177]. AI is not a labelled indication, and the amounts are applied by extrapolation from the prevention of caries and hypersensitivity.

Source: Dutch SmPC of Duraphat® 50 mg/mL, June 2025 revision [177]. The six-monthly frequency coincides with the AAPD recommendation for children at risk of caries [176].

Hypersensitivity prevents brushing and should be treated from the outset and throughout treatment [23]. The British consensus proposes warming the water for the toothbrush, regular rinsing with fluoride mouthwash, desensitising agents such as 10% casein phosphopeptide-amorphous calcium phosphate (CPP-ACP; Tooth Mousse®) and arginine toothpastes, local anaesthesia for scaling, chlorhexidine mouthwash for a limited time before scaling when the gingiva is inflamed, hyperaemic or hyperplastic, and restorations when sensitivity results from thin enamel or breakdown [23]. Reducing sensitivity is a long-term goal [20].

The French PNDS includes a periodontal assessment with clinical indices (plaque and calculus; erythematous, oedematous, smooth and soft gingiva; bleeding on brushing, contact or probing; gingival hypertrophy or hyperplasia; tooth mobility; recessions) and radiographs for bone loss and furcation lesions [19]. The causes of plaque differ with the phenotype: in hypoplastic AI, the rough, irregular surface retains plaque and is associated with frequent gingivitis; in hypomineralised AI, hypersensitivity makes brushing difficult [19]. Management begins with motivation for hygiene and dietary advice, which may form part of a therapeutic education session; in syndromic forms with intellectual disability, hygiene must be ensured by a third person [19]. Prosthetic rehabilitation first requires a stabilised periodontium, sometimes with periodontal surgery, such as crown lengthening to gain retention height, align the necks or correct the gingival architecture; after the definitive prosthesis, periodontal maintenance takes place every 3 or 6 months according to individual risk, and protective night guards may be used because of the large number of restorations [19].

7.3. Primary dentition

In the primary dentition, the aim is to restore function, maintain arch parameters and avoid loss of vertical dimension [22, 24]. The German guideline recommends direct restorations, with or without matrices, on anterior and posterior teeth of the primary and mixed dentitions to relieve pain, restore function and improve aesthetics (recommendation 13, low evidence, strong), and prefabricated stainless steel crowns on primary molars at an early stage to restore shape and occlusion (recommendation 14, low evidence, strong) [24]. ERN CRANIO indicates preformed paediatric crowns between 2 years and the onset of the mixed dentition in cases of severe caries or absent enamel and to prevent or restore loss of vertical dimension: metal (nickel-chromium) on primary molars when there is no aesthetic demand and zirconia on all primary teeth with severe defects and aesthetic concerns; glass ionomer serves as a provisional treatment in very young or poorly cooperative children, and composite is preferred when possible; space maintainers or paediatric prostheses compensate for lost teeth [20]. The German guideline recommends space maintenance and/or provisional prosthetic measures after extraction of a primary tooth (recommendation 7, moderate) [24]. The AAPD describes composite or veneered anterior crowns and full-coverage or veneered posterior steel crowns, and the Orphanet review polycarbonate or composite crowns on anterior teeth [1, 22]. These treatments are often carried out under general anaesthesia or conscious sedation, for behavioural reasons or because of rapid wear; delaying the decision may require later anaesthesia to extract carious and painful teeth [1, 20]. The French PNDS describes the same options (provisional glass ionomer in very young or poorly cooperative children; direct composite, with or without a form, or indirect; preformed paediatric nickel-chromium crowns on primary molars without aesthetic demand and zirconia on all primary teeth with significant defects and aesthetic concerns; space maintainer or paediatric prosthesis) and provides, depending on age, cooperation and complexity, conscious sedation by inhalation of an equimolar mixture of oxygen and nitrous oxide (EMONO), sedation with midazolam or general anaesthesia in hospital [19]. In its follow-up table, between 2 years and the onset of the mixed dentition, it places direct composite restorations with a polycarbonate form or paediatric zirconia crowns on anterior teeth in severe hypoplastic forms, and preformed metal or zirconia crowns on posterior teeth in severe forms where composite is not possible [19].

7.4. Mixed and young permanent dentition: phenotype-based approach

Eruption of the permanent teeth, which extends over about 6 years, is the most difficult period: hypersensitive or crumbling teeth discourage hygiene and are difficult to restore, incomplete gingival maturation limits indirect restorations, the large pulp increases the risk of pulp exposure with crown preparations, and definitive treatment must be compatible with orthodontics and, ideally, with the end of growth [1, 20, 23]. Bonding is a central problem: in hypomineralised AI, the lower mineral content and retained organic matter compromise resin bonding, and the durability of direct composite restorations is much lower in the hypomature and hypomineralised forms than in the hypoplastic form [20, 23]. Pretreatment with 5% sodium hypochlorite, under rubber dam, has been proposed to improve bonding in the hypocalcified form, with weak evidence [23].

The German guideline recommends, for permanent molars, prefabricated steel crowns or individualised single indirect restorations at an early stage (recommendation 15, low evidence, moderate); for permanent anterior teeth, individualised single indirect restorations already in the mixed dentition (recommendation 16, low, moderate); and single indirect crowns in the mixed or early permanent dentition, after eruption and adjustment of the occlusal relationship, as part of a long-term interdisciplinary strategy (recommendation 19, low, strong) [24]. The British consensus proposes flowcharts by phenotype, to be used under local anaesthesia and/or sedation, which start with conservative options and progress according to severity, with reassessment after each intervention [23].

Source: 2026 British consensus [23]. The consensus stresses that mixed phenotypes are frequent, that severe hypomaturation forms behave like hypocalcified ones and that the flowcharts are not intended for treatment plans under general anaesthesia [23]. A non-bonded composite mock-up helps to show the possible result and manage expectations, and layers of composite of different translucency with opaquer help to mask discoloration [23]. Interceptive extractions are considered only case by case and with the orthodontist, because, unlike in MIH or caries, the unerupted teeth are also affected; retaining roots without apical pathology may preserve bone for future rehabilitation [23].

The French PNDS bases caution in the mixed dentition on the immaturity of the permanent tooth at eruption: maturation of the pulp–root complex takes about 1.5 years in incisors and 3 years in molars, the large pulp increases the risk of exposure during crown preparations, and an overly mutilating preparation may cause irreversible pulp inflammation [19]. As a provisional measure, covering the occlusal surfaces of molars and premolars with glass ionomer prevents pain until eruption is complete; preformed nickel-chromium crowns can be placed on first permanent molars without preparation in hypoplastic AI, with prior separation with orthodontic elastics, or with minimal preparation; they preserve vertical dimension, limit attrition and biofilm and had, in a series cited by the PNDS, a failure rate of 4% [19]. Direct composite, preferred on anterior teeth for aesthetic reasons and possible as a provisional restoration of posterior teeth, lasts significantly less in hypomature or hypomineralised AI than in hypoplastic AI, and some authors prefer crowns in the hypomineralised form; the PNDS regards direct composite as an interim solution until the end of growth, whereas indirect bonded composite or ceramic restorations (veneers, crowns, onlays and inlays) have greater longevity regardless of the type of AI, although with few studies; acrylic resin or polycarbonate crowns are economical, require minimal preparation and had satisfactory results at 5–7 years in one clinical report [19].

In adolescence, indirect restorations are justified when prevention (varnishes, fluorides, resin infiltration) and direct restorations do not meet clinical or psychological needs [19]. The French PNDS first calls for an examination of static occlusion (maximum intercuspation, vertical dimension of occlusion, myostabilised centric relation) and an aesthetic analysis of the smile, and recommends preparations limited to removing undercuts, without a peripheral chamfer finish line; this “no-preparation” approach is possible because resin can be polished at the margin, which ceramic does not allow, and it spares enamel for the adult ceramic restorations, which will require more pronounced margins [19]. It prefers individualised pieces, made by conventional methods or by computer-aided design and manufacturing (CAD/CAM, chairside or in the laboratory), to prefabricated steel or polycarbonate veneers or crowns, whose thick cement layer is the weak point; milled composite blocks, with a higher degree of monomer conversion, give a better surface finish and strength, and the bonding joints should be well polished to facilitate cervical hygiene [19]. In forms with anterior open bite, in which contacts are limited to molars and premolars with increased vertical dimension, correction requires collaboration with orthodontics and orthognathic surgery, and rehabilitation of the molars precedes surgery; regular clinical and radiographic follow-up monitors leakage, fractures and loss of occlusal stops, and composite resins may need periodic polishing if they become stained [19].

7.5. Permanent dentition: definitive rehabilitation

The German guideline recommends individualised single indirect crowns, alongside direct restorations, for pain relief, aesthetics and long-term function (recommendation 18, low evidence, strong), starting definitive treatment in the early permanent dentition with the concept of full oral rehabilitation (recommendation 20, low, strong) and definitive single indirect crowns in the permanent dentition (recommendation 21, moderate evidence, strong) [24]. The AAPD accepts veneers, full crowns, implants and fixed or removable prostheses, and an occlusal splint to re-establish vertical dimension when full rehabilitation is required [22]. When the enamel is intact but discoloured, bleaching and microabrasion improve the appearance; on hypocalcified enamel, composite or porcelain veneers may be retained by bonding; if the enamel or dentine does not bond, full-coverage restorations are needed [22].

The strongest evidence on restoration in AI comes from ceramic crowns. In a Swedish randomised split-mouth trial, with patient blinding, 27 adolescents and young adults aged 11 to 22 years received 227 crowns (119 Procera zirconia and 108 IPS e.max Press); after 13.4 ± 1.0 years (11.0–15.3), overall survival was 97.8%, with no difference by material (p = 0.110) or AI subtype (p = 0.974); 97.0% of the surviving crowns had excellent or acceptable quality and 76.1% excellent marginal integrity; sensitivity on a visual analogue scale fell from 4.6 to 0.9 (p < 0.001), gingival bleeding decreased and there were endodontic complications in 3.1% of teeth, mainly with previous trauma [161]. A 2017–2023 review of patients under 19 years with non-syndromic AI (38 of 335 publications) concluded that all forms of restoration reduce pain and sensitivity and improve aesthetics, that composite was used mainly in hypoplastic AI with mild symptoms and that single ceramic crowns have a high success rate in all types of AI and may be the first choice in young people; prosthetic rehabilitation of adolescents with severe AI is cost-effective [84]. A 2024 systematic review included only 4 of 138 studies and found better quality and longevity of ceramic restorations, but called for higher-quality trials [178]. ERN CRANIO advocates, on the basis of these data, earlier intervention with ceramic crowns in adolescence, which is more economical than delaying definitive treatment [20]. The Orphanet review adds an observation relevant to shared decision-making: adults with AI reported that professional restraint in delaying the first invasive restorations had seemed paternalistic to them and that they would have preferred aesthetic restorations earlier, in adolescence [1].

The French PNDS steers adult rehabilitation towards minimally invasive dentistry, which preserves the altered enamel rather than removing it with full peripheral preparations: bonded ceramic veneers, overlays and veneerlays are the approaches to favour, and crowns have had excellent results [19]. When tooth volume is sufficient, indirect restoration is preferred, as it is more durable, and earlier provisional restorations, such as the resins of the mixed dentition, should be removed to standardise the materials and their wear [19]. Planning includes examination of static occlusion and eccentric movements; reconstruction of intercuspation at the reference vertical dimension, or at a therapeutic dimension when there is wear, tested with resin occlusal stops and with verification of the freeway space during swallowing and speech; recording in myostabilised centric relation with a facebow in full-arch rehabilitations; and, in complex cases, a diagnostic wax-up transferred to the mouth as a mock-up, which the patient validates from the aesthetic and occlusal standpoint [19]. Preparations are made through the mock-up, with palatal wrap-around on anterior veneers in the hypomineralised and hypomature forms, overlay or veneerlay designs on posterior teeth and supragingival margins [19].

Bonding depends on the phenotype [19]. On hypoplastic enamel, acid etching produces shallow dissolution and prolonging it adds little; etch-and-rinse systems slightly increase bonding, and universal adhesives used in this mode are considered the option of choice; bond strength values are higher than in the hypomineralised and hypomature forms. In these, the higher protein content justifies deproteinisation with 5% sodium hypochlorite for 60 seconds after etching and rinsing, which increased bonding to hypomineralised enamel, as in MIH and fluorosis; deproteinisation may also form part of the protocol on hypoplastic enamel, but it has no effect on dentine [19]. In severe forms with exposed dentine, the dentine becomes hypermineralised, comparable to sclerotic dentine, acid-resistant and with a thinner hybrid layer; it needs longer etching with orthophosphoric acid, and its thickness increases with the duration of exposure, which favours early treatment [19]. Fixed zirconia restorations have good mechanical properties, low bacterial adhesion and biocompatibility [19].

When there is tooth loss, the PNDS considers a tooth-supported bridge in a bounded space if the two adjacent teeth are already very destroyed or extensively hypoplastic, and implant-supported rehabilitation when only the space is to be treated without preparing the adjacent teeth; in the free-end edentulous situation, it favours the implant, with the removable prosthesis as an alternative [19]. Throughout adult life, patients need maintenance and renewal of restorations, with several successive sets of prostheses; even minimal preparations are performed under local anaesthesia, and in syndromic forms with neurodevelopmental disorder deep sedation with midazolam or general anaesthesia should be considered [19].

7.6. Orthodontics and surgery

The German guideline recommends starting orthodontic or orthodontic-surgical treatment of malocclusion according to the indication (recommendation 17, very low evidence, moderate) [24]. ERN CRANIO recommends an orthodontic examination at 6–8 years; treatment is carried out in phases (impactions, eruption and space problems, sometimes with serial extractions; growth modification with bite-block appliances or face mask in younger children; maxillary expansion in constriction; alignment), with reverse planning with the restorative dentist to leave or redistribute spaces for crowns and re-establish vertical dimension, sometimes with extraction or intrusion of the second molars; retention uses bonded wires and night-time splints [20]. Bracket bonding fails more often, and treatments are longer, with more complications and more pain [20, 23]. Aligners are better accepted by patients with hypersensitivity, but retention is limited by the reduced crown height and lack of undercuts [20]. In adulthood, skeletal discrepancies are reassessed and anterior open bite may require surgical as well as restorative correction [1, 20, 22]. Abnormal oral habits justify assessment by a speech and language therapist [20].

The French PNDS details orthodontics by age [19]. In the primary dentition, care is the responsibility of the paediatric dentist; in the mixed dentition, the orthodontic consultation follows the restorations, assesses the growth pattern and mainly intercepts transverse and vertical anomalies. Dysfunctions are treated at any age: nasal breathing and normal tongue position and mobility, so as not to aggravate the open bite, with oral screens, tongue cribs, nocturnal lingual envelopes or functional trainers, speech therapy and physiotherapy. Before the growth peak, removable appliances with posterior bite-blocks limit compensatory molar extrusion, a Fränkel appliance corrects class II and III early, Hawley plates with a midline screw expand the maxilla or close diastemas, with attention to the weak retention of the abutment teeth, and the quad-helix or Hyrax-type expander are cemented on previously crowned upper first molars [19]. In adolescence, orthodontics prepares spaces for restorations, manages eruption delays and impactions and corrects malocclusions by taking advantage of growth; vertical elastics for open bite are used, but the stability of these compensations during growth is debatable; orthodontic extractions or extractions of non-restorable teeth may be necessary, and aligners are suitable for mild malocclusions [19]. In adults, orthodontics facilitates prosthetic rehabilitation and reduces preparations; there are compromise treatments when surgery is refused; skeletal anchorage, with buccal and palatal corticotomies and vertical intrusion from zygomatic anchors supplemented by a transpalatal arch, treats open bite with hyperdivergence; and orthognathic surgery (Le Fort I with posterior impaction and mandibular rotation, or virtually planned bimaxillary osteotomy with cone-beam tomography) is reserved for cases in which orthodontics is not predictable [19]. Bracket bonding to affected enamel often fails and may cause caries or fractures, including during traction of impacted teeth; if the residual enamel is fragile, prior restorations are placed; on hypomineralised or hypomature enamel, 5% sodium hypochlorite is applied for 1 minute, followed by rinsing and drying, before bonding; and the patient should be informed of the risk of enamel fracture when the brackets are removed [19].

7.7. Behaviour management, anaesthesia and psychological support

The AAPD and the British consensus stress sensitivity to emotional demands, an early relationship with the child and family, behaviour management at each stage and professional psychological support when there are psychosocial consequences [22, 23]. Patients and parents should know from the outset the burden of treatment, the time, possible costs and maintenance, and organising treatment in blocks with realistic goals helps to avoid burnout [23]. Psychological support is particularly important at puberty [20]. General anaesthesia is an option when needs are extensive and treatment is not otherwise possible [1, 24].

The French PNDS adds that quality of life should be assessed with subjective data, in addition to normative dental indices, and takes up the AAPD recommendation of psychological assessment at each stage of treatment, with counselling of the child and parents when there are difficulties; psychological and medico-social support should be offered to the whole family, from the announcement of the diagnosis and throughout follow-up, and contact with patient organisations meets the need for sharing [19]. Person-centred therapeutic patient education begins with an individual educational diagnosis, continues with individual or group sessions by the multidisciplinary team (dentist, orthodontist, geneticist, dietitian, psychologist), with possible participation of the patient organisation and open to parents, and ends with an assessment of the skills acquired; it covers knowledge of the disease, hygiene and diet, the care pathway, what to do in risk situations (gingival inflammation, caries) and complications (debonding of a restoration or a bracket) and psychological aspects [19]. The O-Rares centre in Strasbourg declared to the regional health agency in 2021 a specific programme for rare diseases of the teeth and oral cavity (DentO-RarEduc) [19]. Each treatment has its own contraindications, to be discussed with the patient and the team, and prosthetic treatments are based on a written plan with a cost estimate [19].

The PNDS mentions sedation with EMONO or midazolam without stating doses; the official labels define them for procedural sedation or analgesia, and they are applied to dental care in AI by extrapolation [19, 179, 180].

Sources: Spanish SmPCs of Ozalin® (March 2026 revision) and Kalinox® (June 2024 revision) [179, 180].

7.8. Treatment of syndromic forms

Enamel-renal syndrome. There is no disease-modifying treatment; management is multidisciplinary, with a specialist dental team for restorations, impacted teeth and gingival hyperplasia, renal imaging and long-term nephrological follow-up, and a genetics consultation [5, 6, 115]. In the Lyon series, the four children were treated with oral hyperhydration, dietary advice (sodium- and protein-controlled diet, with age-appropriate calcium intake) and correction of 25-hydroxyvitamin D deficiency with low daily doses of native vitamin D; citrate was not used, because renal phosphate wasting coexisted with profound hypocalciuria, and thiazides were not indicated in the absence of hypercalciuria [116]. The authors raised the hypothesis of a role for burosumab, an anti-FGF-23 antibody, given the hypophosphataemia with raised FGF-23; this is a proposal for discussion from a case series, without a trial, and the drug has no indication in this disease [5, 116]. The same authors recommend general bone-protective measures for all children with enamel problems (age-appropriate calcium intake and native vitamin D supplementation, according to the 2022 French paediatric recommendations) [116].

AI with distal renal tubular acidosis (WDR72). Treatment of dRTA is oral alkali therapy, with bicarbonate and/or citrate salts, usually potassium salts, because hypokalaemia is common; sodium salts are avoided because they aggravate hypercalciuria [5, 14]. The aim is to correct acidosis and hypokalaemia, because compensated acidosis (normal pH with low bicarbonate) is not sufficient for growth; alkali and citrate prevent progression of nephrocalcinosis but do not reverse it, and they reduce the risk of stones [14]. One millimole of citrate is converted in the liver to two millimoles of bicarbonate, and a European survey identified more than 30 alkali formulations in use [14]. GeneReviews is the reference for doses, and the summary of product characteristics (SmPC) of Sibnayal®, approved in the European Union, serves for comparison [14, 181].

SD, standard deviation; ADV7103 is the development name of Sibnayal® [14].

Sibnayal® is a prolonged-release granule formulation of potassium citrate and potassium hydrogen carbonate, in 8 mEq sachets (282 mg potassium citrate and 527 mg potassium hydrogen carbonate) and 24 mEq sachets, indicated for the treatment of dRTA in adults, adolescents and children from 1 year [181]. It is contraindicated in hypersensitivity, glomerular filtration rate of 44 mL/min/1.73 m² or less and hyperkalaemia; between 45 and 59 mL/min/1.73 m² it should only be used if the benefit outweighs the risk, with monitoring of potassium and renal function at least twice a year; caution is required with other drugs that raise potassium or predispose to arrhythmias [181]. The most frequent adverse effects are abdominal pain (14%), upper abdominal pain (8%), gastrointestinal pain (2%) and nausea (2%), as well as diarrhoea, dyspepsia and vomiting, generally mild to moderate and transient, without the need for discontinuation [181]. Twice-daily dosing facilitates adherence at school age [14]. Potassium-sparing diuretics should be used with caution or avoided [5]. The AI itself is treated by the dentist, as in isolated forms [14]. Treatment and surveillance of dRTA follow the recommendations for hereditary renal tubular acidoses.

Junctional epidermolysis bullosa. Dental care is necessary because of the inherent enamel anomalies [15]. For the wounds of junctional and dystrophic epidermolysis bullosa (EB), from 6 months of age, birch triterpenes gel (Filsuvez®) is approved, which in the phase III trial with 223 patients achieved complete closure of the target wound within 45 days in 41.3% versus 28.9% with the control gel (relative risk 1.44; 95% CI 1.01–2.05), with no difference at 90 days; the gene therapies approved for EB are specific to COL7A1 and do not apply to the laminin-332 genes or to COL17A1 [5]. None of these treatments acts on the enamel.

Tricho-dento-osseous syndrome. There is no targeted treatment; the published guidance is a review proposal, without trials, based on a team with paediatric dentistry, orthodontics, endodontics, oral surgery, prosthodontics and radiology, together with paediatrics and genetics, baseline documentation (intraoral radiographs, panoramic radiograph, lateral cephalometric radiograph, models and photographs) and a short-term plan with long-term follow-up [128]. The aims are to prevent hypersensitivity, caries, abscesses and loss of vertical dimension due to attrition, restore function and improve aesthetics [128]. For mild to moderate hypersensitivity, dietary advice (fewer acidic foods, water or milk after acidic drinks, use of a straw), a soft toothbrush and desensitising toothpastes, mouthwashes or lozenges with potassium nitrate or sodium fluoride, casein phosphopeptide-amorphous calcium phosphate (CPP-ACP) paste and toothpaste with 8% arginine and calcium carbonate are proposed; for severe hypersensitivity, in-office high-concentration fluoride varnish and, if these fail, dentine adhesives, glass ionomers or composite resin to occlude the tubules, bioglass, laser or calcium silicate cements [128]. On molars, preformed stainless steel crowns are placed as soon as the tooth reaches occlusal height, as an intermediate measure that can be replaced later; to spare tooth structure, glass ionomer or resin is placed on the occlusal surface of the erupting molar so that occlusal reduction takes place in the material, the teeth are separated beforehand with elastics and the crown is cemented with glass ionomer, a material also preferred for provisional restorations, whereas amalgam is not advised [128]. Full ceramic crowns on anterior teeth are not recommended before 16 years, because of the large pulps and immature gingival contour; until then, direct composite resin veneers or ceramic veneers without preparation or with minimal preparation are used [128]. Endodontic treatment is difficult: the elongated pulp chamber makes it difficult to locate, clean and obturate the canals, and there are frequently pulp stones and atypical apical systems; the combination of apical lateral condensation with vertical condensation in the chamber, the mineral trioxide aggregate (MTA) apical plug in immature teeth and vital pulpotomy as an alternative to pulpectomy in hypertaurodontism have been described [128]. In a series cited in that review, endodontic treatment of first permanent molars in children aged 8 to 16 years had a 36% success rate, a figure obtained outside TDO and attributed to thin walls, open apices and wide canals [128]. Taurodont molars are weak prosthetic abutments, because of the smaller inserted root surface and the absence of cervical constriction, and posts should be avoided; extraction often requires elevators, because of the apical position of the furcation; when the first permanent molars have a poor prognosis, extraction is planned with orthodontics at the onset of calcification of the bifurcation of the second molars, first the lower and six months later the upper molars [128]. Anxiety is managed with local anaesthesia, nitrous oxide sedation or general anaesthesia, and rubber dam protects sensitive teeth from air and water [128]. The bone changes have no specific treatment, and the sources consulted do not describe implant outcomes in TDO; the absence of alveolar ridge resorption after tooth loss, described in one family, is favourable to prosthetic rehabilitation [149, 151]. In the French family with attenuated TDO, recurrent infections led to extraction of the teeth and complete removable dentures in the adolescent son and in the mother's upper jaw [123].

Jalili syndrome. Management is symptomatic and shared between ophthalmology (low vision support, red-tinted filters recommended in the original description of the syndrome, monitoring of progression) and dentistry [5, 154].

Kohlschütter-Tönz syndrome. Seizures are treated with conventional antiepileptic drugs and are often refractory; in the SLC13A5 form, in a prospective natural history study with 30 patients, valproic acid was the most frequently used drug (22 patients) and 80% of caregivers considered it useful, observational data reported by caregivers; in the ROGDI form, perampanel has been described as effective in isolated cases, off-label [5]. Dental care is a substantial part of management, because of hypersensitivity and the high caries risk, and treatment under general anaesthesia is often the feasible route [5]. Adeno-associated virus gene therapy for SLC13A5-related disease is at the preclinical stage [5]. In ROGDI patients, perampanel was associated with control of super-refractory status epilepticus in a Latvian child and, in a Chinese child, with about seven months without seizures and gross motor improvement, in combination with levetiracetam, after recurrence on valproic acid; these are isolated reports, without a trial, and the use is off-label [141, 156]. In the SLC13A5 cohort, higher doses of valproic acid correlated with the benefit reported by caregivers [155].

Dental anomalies with short stature (LTBP3). In addition to specialist dental and orthodontic care, the authors of a series of three siblings recommend that cardiologists actively look for valvular and aortic root abnormalities even in the absence of symptoms [5, 157].

Heimler syndrome and Zellweger spectrum. Dental treatment of the AI forms part of the multidisciplinary care of the Zellweger spectrum, with dental review every 6 months [5].

CRAC channelopathies (ORAI1, STIM1). Haematopoietic progenitor cell transplantation corrects the immunodeficiency but not the enamel defect, the myopathy or the anhidrosis: teeth that erupt after a successful transplant still have defective enamel and require extensive dental treatment, which should be explained to the family before transplantation [5, 133].

7.9. Regulatory status

7.10. Clinical trials and studies

There are no drug trials in AI. The only randomised trial with long follow-up compares two ceramic materials in a split-mouth design and does not compare crowns with other strategies; the remaining studies are observational or small, and the registrations with unknown status have not published results [161, 178, 182]. Indirect comparisons between restorative series are not valid, because they differ in age, phenotype, severity and success criteria.

7.11. Developments in the last 12 months

Other approaches remain preclinical: chemical chaperones such as 4-phenylbutyrate, proposed for forms with endoplasmic reticulum stress, and biomimetic enamel repair materials, whose efficacy in AI has not been demonstrated [6]. No marketing authorisation application or drug trial for AI was identified in the searches carried out, which does not prove that none exist.

7.12. Special populations and situations

In young children, behaviour management and general anaesthesia or sedation are often necessary, and preformed crowns and fluoride varnish are the core measures [20, 24]. Transition to adult care is planned by the paediatric team, which should hand over to the restorative dentist a motivated patient, with good hygiene and without compromised treatment options [1, 23]. The sources consulted do not describe specific recommendations for pregnancy or for the perioperative period in isolated AI; in syndromic forms, those of the organ involved apply (for example, acid–base control in dRTA). The French PNDS places transition between 15 and 19 years, depending on the patient, with detailed reports between specialists; the first adult appointment revisits the history and examination, reinforces the importance of regular follow-up and of situations requiring urgent consultation, provides information on practices to avoid and social rights, assesses understanding of the disease, transfers the radiographs to avoid repetition, re-explains transmission, assesses the risk of recurrence in a pregnancy and adapts the plan; adults who have not been diagnosed, or who are treated without links to a centre, benefit from an assessment appointment at an O-Rares centre [19]. In an emergency, professionals should be aware of the diagnosis of AI, the associated malformations and involvement and the treatments under way, and the health booklet or electronic record is the linking instrument [19].

7.13. Convergences and divergences between documents

The documents agree on the need for early diagnosis and intervention, intensive prevention, a multidisciplinary approach and psychological attention [20, 22–24]. They diverge on the timing and type of restoration: the German guideline recommends single indirect restorations on permanent anterior teeth already in the mixed dentition and the start of full rehabilitation in the early permanent dentition, and ERN CRANIO advocates ceramic crowns from early adolescence; the British consensus favours minimal intervention, considers preformed metal crowns only provisional and warns about indirect restorations before gingival maturation; the 2013 AAPD guideline does not grade recommendations [20, 22–24]. The divergence reflects the scarcity of comparative trials and the differences between health systems.

The French PNDS converges with the other documents on principles and adds its own divergences [19]. On dental follow-up, the PNDS summary calls for check-ups at least twice a year, whereas the text sets six-monthly follow-up before 6 years and at least annual follow-up thereafter, if the disease is stable, with six-monthly check-ups in children with prosthetic restorations and at least six-monthly in the rehabilitated adult; ERN CRANIO and the AAPD speak of regular examinations without a fixed interval [19, 20, 22]. The PNDS accepts preformed nickel-chromium crowns on first permanent molars, without preparation, with a low failure rate and as an interim measure until the end of growth; the British consensus considers them only provisional, and the German guideline places them on a par with individualised indirect restorations [19, 23, 24]. The PNDS applies 5% sodium hypochlorite after acid etching and rinsing in ceramic bonding to enamel, and before bracket bonding, whereas the British consensus describes it as a pretreatment, with weak evidence in both [19, 23]. The PNDS 1000 ppm toothpaste lies at the lower limit of the AAPD range (at least 1000 and less than 1500 ppm), which accepts 5000 ppm prescription toothpastes [19, 176]. The first orthodontic assessment is at 6 years in the PNDS and at 6–8 years in ERN CRANIO [19, 20]. Like the AAPD, the PNDS does not grade its recommendations, which are based on case series and the consensus of the O-Rares network experts [19].

8. Reproductive Options and Family Counselling

8.1. Risks by mode of inheritance

Counselling begins with determining the mode of inheritance, which patients and families find useful even before molecular diagnosis, and which is as important as, or more important than, the phenotype for discussing risks [1]. Genetic heterogeneity makes empirical estimation fallible when the gene is not known, and molecular diagnosis gives precision to counselling [1, 2].

The French PNDS adds that an adult with recessive AI has a very low risk of having affected children, except in a union with a person affected by the same disease or in consanguinity, and that heterozygosity testing may be offered to the family [19].

8.2. Cascade testing and relatives

Once the variant has been identified, the segregation study in relatives confirms its relationship with the disease and allows mildly affected individuals and carriers to be identified, as the Strasbourg laboratory does for class 4 and 5 variants [2]. In genes with two inheritance patterns, testing relatives has different implications: in the family of a patient with JEB, carriers of LAMB3, LAMA3 or COL17A1 may benefit from a dental examination, and in the family of a patient with dominant AI due to LAMB3, the last-exon variant does not predict JEB [5, 6, 15]. In syndromic forms, the molecular diagnosis guides the surveillance of affected relatives (kidney, eye, heart, hearing) [2, 5]. The French PNDS recommends trio analysis from the outset; after diagnosis in the index case, testing may be extended to symptomatic relatives and to adult relatives at risk of carrying one or two variants, in accredited laboratories [19].

In enamel-renal and Kohlschütter-Tönz syndromes, inheritance is autosomal recessive: the parents of an affected individual are generally heterozygous carriers without manifestations, and the risk for each sibling is 25% [5, 146]. Homozygosity does not prove that both parents are carriers: in a Chinese child homozygous for a ROGDI variant, only the mother was a carrier, which led to suspicion of uniparental disomy of chromosome 16, not confirmed in the report, and testing both parents is necessary to estimate the recurrence risk [141]. When there is a founder variant, such as the FAM20A variant in the Brazilian families, carrier testing may be extended to relatives and to spouses of the same origin [137].

In TDO and AI4, inheritance is autosomal dominant with high penetrance, and each child of a carrier has a 50% probability of inheriting the variant [18, 128]. Expressivity varies within the same family, from AI with mild taurodontism to complete TDO, so a mildly affected parent may have children with a more marked picture; affected relatives may have no teeth or be edentulous after successive extractions, and the hair phenotype can only be verified in childhood photographs [128, 153]. De novo variants and patients without a family history have been described, which requires testing both parents before estimating the risk for siblings [122, 124]. In the French family, after the molecular diagnosis, testing of carriers and offspring and clinical reassessment of a brother over time were proposed, as his age did not allow the phenotype to be excluded [123].

8.3. Psychological and reproductive aspects

These conditions are embarrassing and distressing and are associated with social exclusion and ridicule, so the interview should be sensitive and support early; affected parents may themselves have experienced the burden of treatment, which influences their expectations [1, 23]. The sources consulted do not describe demand for prenatal or preimplantation diagnosis in AI, nor guidance on testing asymptomatic minors. ERN CRANIO places diagnostic confirmation by a rare disease specialist or geneticist from 2 years of age, with genetic testing especially when a syndrome is suspected [20].

The French PNDS organises the disclosure of the clinical and genetic diagnosis in a dedicated appointment, in a specialist centre or with professionals linked to it; when the person who ordered the test is not a geneticist, disclosure should be made jointly with a geneticist or, failing that, with a genetics appointment scheduled promptly, and the presence of a psychologist is strongly recommended [19]. The appointment conveys the characteristics of the disease, the mode of transmission and genetic counselling, the update of the record and the investigations to look for extraoral anomalies not yet studied, the treatment and follow-up plan by age, the professionals involved, attention to the first reactions, contacts for patient organisations and social steps; this is followed by a summary letter to the dentist, the general practitioner, the corresponding specialists and the family and, on request, a second appointment a few weeks later to go over the information again and introduce the team [19].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence

Prevalence estimates vary about twenty-fold, and each study measures different things [1, 5].

The relative frequency of the genes, based on published families, is presented in section 2.7 and is not equivalent to prevalence [6]. In clinical cohorts, the distribution of phenotypes was 53% hypoplastic, 27% hypomature, 16% hypomineralised and 4% with taurodontism in France, and 70% hypomineralised and 30% hypoplastic in a British sample of 30 children [2, 142].

Syndromic forms are ultra-rare: enamel-renal syndrome has an estimated prevalence below 1 per 1,000,000, with 69 patients published up to 2021; Kohlschütter-Tönz syndrome, brachyolmia with AI (15 cases or families recorded by Orphanet) and Raine syndrome (40 cases or families) are in the same prevalence class [5, 116, 146]. In the Gaza Strip, cone-rod dystrophy has a prevalence of 1:10,000 and Jalili syndrome accounts for 83% of these cases, a figure from an endogamous population that should not be extrapolated [5]. Junctional epidermolysis bullosa has a point prevalence of 2.1 per million in the Netherlands (Dutch registry) [5].

9.2. Founder populations and recurrent variants

The Asian community of Bradford, with frequent consanguinity, contributes to the predominance of Pakistani families in the Leeds cohort, which biases the relative frequencies of recessive genes [108].

TDO has been described mainly in large families: the original family of 107 individuals and the Virginia and North Carolina families with the 4 bp deletion, in which clinical variability did not result from allelic heterogeneity [120, 128]; Finnish families with homeodomain variants [121]; a Swiss family and Korean, Australian, Caucasian and Turkish families with the 2 bp deletion [6, 85, 153]; British families, one of them of mixed Afro-Caribbean and Caucasian background [124]; a Chinese patient with a de novo variant [122]; three families from Taiwan [125]; and a family of African origin followed in France [123]. No Portuguese or Brazilian families with TDO were identified in the sources consulted.

9.3. Brazil and Portugal

In Brazil, the AI LOVD records a Brazilian family with the FAM20A variant c.406C>T p.(Arg136Ter), and in 2026 the first Brazilian families with FAM83H-related AI were described, followed at the Oral Care Centre for Hereditary Diseases of the University Hospital of Brasília [6, 101]. In 2019, five Brazilian families with enamel-renal syndrome (11 patients) were described, studied by the Federal University of the Jequitinhonha and Mucuri Valleys and the State University of Campinas, all homozygous for the founder variant FAM20A c.1447del p.(Glu483LysfsTer24), the largest series from a single population [137]. The 2021 systematic review of enamel-renal syndrome was carried out by authors from Minas Gerais and Campinas [146]. The sources consulted contain no epidemiological data from Portugal, which does not demonstrate that none exist.

9.4. European networks and patient organisations

ERN CRANIO has included, since 2017, centres specialised in rare oral diseases, with an orodental group that published the AI management timeline and is creating a European registry (https://www.ern-cranio.eu; https://www.ern-cranio.eu/ern-cranio-registry), in addition to the clinical consultation platform between members (CPMS 2.0) [20]. In France, the O-Rares reference centres (https://www.o-rares.com), part of the TETECOU network (https://www.tete-cou.fr), have operated since 2016; in the United Kingdom, the UK AI/DI National Clinical Excellence Network brings together paediatric dentists with an interest in AI and dentinogenesis imperfecta; in the Netherlands and Denmark, treatment of rare dental diseases is centralised [20, 23]. The French patient organisation Amélogenèse France can be reached at https://amelogenesefrance.wixsite.com/amelogenese [20]. The sources consulted do not identify patient organisations in Portugal or Brazil.

The French PNDS describes the organisation: the coordinating reference centre in Strasbourg (Hôpitaux Universitaires de Strasbourg) and the constituent site at Rothschild Hospital (Assistance Publique-Hôpitaux de Paris), with 16 competence centres (Angoulême, Besançon, Bordeaux, Dijon, Lyon, Marseille, Montpellier, Nancy, Nantes, Créteil, Pitié-Salpêtrière, Reims, Rennes, Rouen, Toulouse and Tours) [19]. Follow-up is coordinated by a specialist from these centres, with the paediatrician or general practitioner and the patient's dentist, whose role is to confirm the diagnosis at a centre, take part in coordination, provide local care and refer according to needs; in syndromic forms, the AnDDI-Rares (developmental anomalies) and OSCAR (rare diseases of bone, calcium and cartilage) networks are also involved, and an annual appointment at one of these centres coordinates care [19]. The team includes dentists with various competences, a geneticist, radiologist, maxillofacial surgeon, paediatrician or family doctor, specialists in the organs affected, genetic counsellor, speech and language therapist, physiotherapist, psychologist and social worker, with a multidisciplinary consultation meeting when necessary; the maternal and child health doctor and the school doctor help to resolve the child's medico-social difficulties [19].

Financially, the PNDS describes high costs for families: some dental procedures in the fee schedule carry fee supplements (onlays, crowns on permanent teeth), others do not exist in the schedule (onlays on primary teeth) and are paid by the family, reimbursement of orthodontics ends at 16 years, there is no specific coverage of orthodontics in rare dental diseases, unlike cleft palate, and adult orthodontics is not covered [19]. AI is not among the listed long-term conditions (ALD30), but an ALD31 (off-list condition) may be requested, which, when accepted, covers at 100% only reimbursable procedures; benefits from the departmental centres for persons with disabilities rarely apply, exceptional assistance from the health insurance funds depends on resources and on each committee, with no national rule, and transport to appointments at the centres is covered [19].

10. Illustrative Clinical Case

10.1. Hypocalcified AI due to FAM83H with impacted teeth and pre-eruptive coronal resorption

An 18-year-old Brazilian woman, the daughter of non-consanguineous parents with no enamel changes, with previous diagnoses of attention deficit hyperactivity disorder and asthma, was referred to the Oral Care Centre for Hereditary Diseases of the University Hospital of Brasília for delayed eruption and suspected AI [101]. Examination showed facial dysmorphism, anterior open bite, prolonged retention of primary teeth and delayed eruption of the permanent teeth; the teeth had yellow-brown discoloration, surface irregularities and altered morphology, compatible with hypocalcified AI, and many had been restored with composite; orthodontic mini-implants had been placed to extrude impacted teeth, without success, and were removed [101]. The panoramic radiograph showed reduced enamel radiodensity, all permanent teeth present and eleven impacted teeth; cone-beam computed tomography showed enlarged dental follicles and pre-eruptive coronal resorption in nine of the impacted teeth (82%) [101]. Serial panoramic radiographs at 10, 14 and 18 years documented a lower right molar with a normal crown at 10 years and progressive coronal resorption thereafter, up to extensive crown destruction incompatible with caries after eruption; after gingivectomy in the region of the upper right lateral incisor and removal of the composite, the surface of the incisors was rough, which was attributed to pre-eruptive resorption [101]. The exome identified the heterozygous variant NM_198488.5:c.1055C>A p.(Ser352Ter) in exon 5 of FAM83H, classified as likely pathogenic (PVS1 and PM2 criteria), previously described in another family [87, 101]. The parents had no enamel changes and were not tested; the authors consider the event apparently de novo, given the absence of reported incomplete penetrance in this form [101].

In the same article, a second, unrelated 15-year-old girl, also the daughter of unaffected non-consanguineous parents, had anterior open bite, mixed dentition, yellow-brown discoloration and reduced mineralisation of all teeth, poor hygiene with generalised gingivitis, eleven impacted teeth, seven of which (63%) had pre-eruptive coronal resorption, and progressive resorption documented on radiographs at 5, 10 and 15 years; she had the heterozygous variant c.1379G>A p.(Trp460Ter), pathogenic and previously described in five families [101].

The figures actually consulted show the pedigree, the intraoral photographs and the panoramic radiograph with the impacted teeth (https://pmc.ncbi.nlm.nih.gov/articles/PMC13139228/figure/Fig1/), the tomography slices with pre-eruptive coronal resorption (https://pmc.ncbi.nlm.nih.gov/articles/PMC13139228/figure/Fig2/) and the longitudinal radiographic course (https://pmc.ncbi.nlm.nih.gov/articles/PMC13139228/figure/Fig3/) [101].

The case illustrates several points of this article: the dominant hypocalcified phenotype typical of FAM83H, with a truncating variant in the last exon; the possibility of sporadic cases due to de novo mutation in a dominant form; associated anomalies (open bite, impactions, enlarged follicles) listed by the British consensus and ERN CRANIO; and the value of longitudinal radiographic follow-up and cone-beam tomography [20, 23, 101]. Orthodontic traction with mini-implants failed in this case and should not be read as a routine approach; the guidelines recommend joint orthodontic planning with the restorative dentist and treatment according to the indication [20, 24, 101].

10.2. Attenuated TDO due to DLX3 in a family with apparently isolated AI

A 10-year-old boy of African origin was seen in the emergency department of a Paris hospital with cellulitis associated with necrosis of the permanent upper right central incisor [123]. He had no medical history and no changes in hair, skin or nails, according to the parents and clinicians. The enamel was dark yellow, striated, heavily worn and translucent in the worn areas, with radiopacity similar to that of dentine; the panoramic radiograph showed taurodontism of the first permanent molars, short roots, retained and impacted teeth (the first permanent molars and the lower left first primary molar) and diastemas (figure 1 of the article) [123]. The clinical diagnosis was hypoplastic and hypocalcified AI by Witkop's criteria. The mother had orange-brown enamel with loss of more than half of the tooth structure and multifocal high-density mandibular areas, interpreted as florid cemento-osseous dysplasia; bone scintigraphy showed only one focus of hyperossification in the right ankle (figure 2) [123]. Targeted sequencing of a panel of 19 enamel anomaly genes identified in both the heterozygous variant NM_005220.3:c.398G>C p.(Arg133Pro), at the start of the homeodomain, absent in the father and previously described in a Finnish family with typical TDO; it was classified as pathogenic (class 5) according to the recommendations of the American College of Medical Genetics and Genomics, and the diagnosis was revised to attenuated TDO (figure 3, pedigree and alignment) [87, 121, 123]. The article describes the variant as a frameshift, but it is a missense substitution. After several dental infections, the boy underwent extractions and rehabilitation with a complete removable denture, and the mother extraction of the remaining teeth, except two lower molars, with an upper complete and lower partial denture [123]. The case shows that, without genetic testing, a TDO with minimal extradental expression is classified as isolated AI or as another dental dysplasia, with consequences for counselling of the offspring; extraction of the entire dentition in childhood resulted from the severity of the infections and does not correspond to the conservative approach proposed by the guidance documents.

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Isolated AI does not reduce life expectancy, but it is a chronic condition with a lifelong treatment burden, an impact on quality of life and high costs [20, 23]. Without treatment, the enamel deteriorates mechanically, with hypersensitivity and loss of structure and vertical dimension; in TDO, the teeth are lost by the second or third decade [1, 120]. With restorative treatment, pain and sensitivity decrease and aesthetics improve, and ceramic crowns placed in adolescence and early adulthood had a survival of 97.8% after 13.4 years, with no difference between AI subtypes [84, 161]. Direct composite restorations last less in hypomineralised than in hypoplastic forms [20]. The prognosis of syndromic forms is that of the syndrome: in enamel-renal syndrome, the risk is renal (stones, nephrocalcinosis, possible chronic kidney disease if not detected) [116]; in WDR72-related dRTA, growth and renal function depend on control of the acidosis [14]; in Jalili syndrome, the retinal dystrophy is progressive [154]; in Kohlschütter-Tönz syndrome, epilepsy and delay dominate the prognosis [5]. In SLC13A5-related disease, seizures decrease after the first decade, but developmental impairment is global and skills plateau in adolescence [140, 155]. In TDO, intellect is normal, life expectancy has not been described as reduced and bone thickening does not increase fractures; the 2012 review considers that early, age-adapted prevention allows the dentition to be retained in the long term, an assessment not based on series with follow-up [128, 149].

11.2. Surveillance

In the French PNDS, the age-based pathway begins with the first clinical and radiological examination by the patient's dentist between 6 months and 6 years, with referral to a centre; confirmation of the diagnosis and the search for a syndrome, with a geneticist if necessary, take place from 2 years; the initial orthodontic assessment is at 6 years; the first restorations of the permanent teeth (first molars and incisors) are carried out in the stable mixed dentition; and in adults at least one annual appointment to check the stability of the results is planned [19].

12. Summary and Concluding Remarks

Amelogenesis imperfecta is a class of genetic disorders of enamel development that affects all or nearly all teeth of both dentitions without following the chronology of tooth formation. Witkop's 1988 classification, with four phenotypic types (hypoplastic, hypomaturation, hypocalcified and hypomaturation-hypoplastic with taurodontism) and fifteen subtypes also defined by inheritance, remains the common language of the clinic, and it was updated in 2023 with the known genes and a syndromic class; the hypocalcified and hypomature phenotypes are now seen as a spectrum of hypomineralisation. More than 70 genes are involved, with mechanisms ranging from matrix proteins (AMELX, ENAM, AMBN) to proteases (MMP20, KLK4), adhesion (LAMB3, ITGB6, FAM83H), transport and pH (WDR72, SLC24A4, GPR68) and developmental regulators (FAM20A, DLX3); several genes have dominant effects through truncated proteins that escape NMD, and some show both dominant and recessive inheritance. In published families, FAM83H, FAM20A, ENAM and AMELX account for about 60% of solved cases, and current panels achieve a diagnosis in about 60% of patients.

Diagnosis rests on clinical examination, the pedigree and radiographs, and genetic testing is now recommended when the clinical picture suggests a genetic cause, with copy number variation analysis and attention to nomenclature, because the reference variant database contains outdated notations. Molecular diagnosis changes follow-up whenever it reveals a syndromic form: enamel-renal syndrome requires renal surveillance, WDR72-related AI acid–base assessment, Jalili syndrome retinal assessment, LTBP3-related dental anomalies cardiac assessment, and Heimler syndrome hearing assessment. The main differentials are fluorosis and molar-incisor hypomineralisation, distinguished by distribution, symmetry and family history.

Treatment is dental and begins at eruption of the first tooth, with intensive prevention, fluoride varnish and control of hypersensitivity, preformed crowns and direct restorations in the primary dentition, phenotype-adapted restorations in the mixed dentition and definitive rehabilitation, often with ceramic crowns, from adolescence; the German S3 guideline, the British consensus, ERN CRANIO, the French PNDS and the AAPD converge on principles and diverge on the timing of indirect restorations. There are no drugs directed at AI; in syndromic forms, the organ involved is treated. The psychosocial burden is high from childhood, and psychological support, shared decision-making and access to specialist centres of the European networks are part of care.

13. References

1. Crawford PJ, Aldred M, Bloch-Zupan A. Amelogenesis imperfecta. Orphanet J Rare Dis. 2007;2:17. doi:10.1186/1750-1172-2-17
2. Bloch-Zupan A, Rey T, Jimenez-Armijo A, Kawczynski M, Kharouf N, O-Rare consortium, et al. Amelogenesis imperfecta: next-generation sequencing sheds light on Witkop's classification. Front Physiol. 2023;14:1130175. doi:10.3389/fphys.2023.1130175
3. Orphanet. Amelogenesis imperfecta (ORPHA:88661). Available from: https://www.orpha.net/en/disease/detail/88661
4. Online Mendelian Inheritance in Man (OMIM). Morbid map (morbidmap.txt, local version of August 2026). Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/downloads
5. GeneGuidance. Gene records for AMELX, ENAM, AMBN, FAM83H, MMP20, KLK4, WDR72, SLC24A4, LAMB3, LAMA3, COL17A1, FAM20A, ITGB6, AMTN, GPR68, RELT, DLX3, CNNM4, LTBP3, ROGDI, SLC13A5, FAM20C, PEX1, PEX6, AIRE, STIM1, ORAI1 (version of 2026-09-26).
6. Smith CEL, Poulter JA, Antanaviciute A, Kirkham J, Brookes SJ, Inglehearn CF, et al. Amelogenesis imperfecta; genes, proteins, and pathways. Front Physiol. 2017;8:435. doi:10.3389/fphys.2017.00435
7. Hany U, Watson CM, Liu L, Nikolopoulos G, Smith CEL, Poulter JA, et al. Novel ameloblastin variants, contrasting amelogenesis imperfecta phenotypes. J Dent Res. 2024;103(1):22-30. doi:10.1177/00220345231203694
8. Orphanet. Hypoplastic amelogenesis imperfecta (ORPHA:100031). Available from: https://www.orpha.net/en/disease/detail/100031
9. Orphanet. Hypocalcified amelogenesis imperfecta (ORPHA:100032). Available from: https://www.orpha.net/en/disease/detail/100032
10. Orphanet. Hypomaturation amelogenesis imperfecta (ORPHA:100033). Available from: https://www.orpha.net/en/disease/detail/100033
11. Orphanet. Hypomaturation-hypoplastic amelogenesis imperfecta with taurodontism (ORPHA:100034). Available from: https://www.orpha.net/en/disease/detail/100034
12. Orphanet. Amelogenesis imperfecta-gingival hyperplasia syndrome (ORPHA:171836). Available from: https://www.orpha.net/en/disease/detail/171836
13. Orphanet. Trichodysplasia-amelogenesis imperfecta syndrome (ORPHA:79129). Available from: https://www.orpha.net/en/disease/detail/79129
14. Alexander RT, Gil-Peña H, Greenbaum LA, Santos F. Hereditary distal renal tubular acidosis. 2019 Oct 10 [updated 2025 Apr 3]. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK547595/
15. Pfendner EG, Lucky AW. Junctional epidermolysis bullosa. 2008 Feb 22 [updated 2018 Dec 20]. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1125/
16. Bäckman B, Holm AK. Amelogenesis imperfecta: prevalence and incidence in a northern Swedish county. Community Dent Oral Epidemiol. 1986;14(1):43-7. doi:10.1111/j.1600-0528.1986.tb01493.x
17. Witkop CJ. Hereditary defects in enamel and dentin. Acta Genet Stat Med. 1957;7(1):236-9. doi:10.1159/000150974
18. Orphanet. Tricho-dento-osseous syndrome (ORPHA:3352). Available from: https://www.orpha.net/en/disease/detail/3352
19. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Amélogenèses imparfaites. Saint-Denis La Plaine: HAS; 2021 (text dated August 2021, published September 2021). Available from: https://www.has-sante.fr/jcms/p_3284536
20. Jimenez-Armijo A, Carvajal Monroy PL, Ombashi S, Fennis WMM, Ross JN, Van Den Boogaard MH, et al. Timeline of amelogenesis imperfecta management. J Craniofac Surg. 2025;36(8):2921-5. doi:10.1097/SCS.0000000000011344
21. Martin M, Gowda S, Foster Page L, Thomson WM. Oral health-related quality of life in Northland Māori children and adolescents with Polynesian amelogenesis imperfecta. Front Dent Med. 2024;5:1485419. doi:10.3389/fdmed.2024.1485419
22. American Academy of Pediatric Dentistry. Guideline on dental management of heritable dental developmental anomalies. Pediatr Dent. 2016;38(6):302-7 (Reference Manual 2016-17; 2013 revision). Available from: https://www.aapd.org/assets/1/7/G_OHCHeritable2.PDF
23. Lakhani S, Monteiro J, Agel M, Lyne A, Somani C, Balmer R, et al. A UK-based consensus on clinical decision flowcharts for managing childhood amelogenesis imperfecta in the permanent dentition. Eur Arch Paediatr Dent. 2026;27(1):135-45. doi:10.1007/s40368-025-01127-1
24. Dujic H, Bücher K, Schüler IM, Schmidt P, Hertel S, Timpel J, et al. The German S3-level guideline for the diagnosis and management of rare dental disorders. Clin Oral Investig. 2026;30:459. doi:10.1007/s00784-026-07142-w
25. Online Mendelian Inheritance in Man (OMIM). #104530 AMELOGENESIS IMPERFECTA, TYPE IA; AI1A. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/104530
26. Online Mendelian Inheritance in Man (OMIM). #104500 AMELOGENESIS IMPERFECTA, TYPE IB; AI1B. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/104500
27. Online Mendelian Inheritance in Man (OMIM). #204650 AMELOGENESIS IMPERFECTA, TYPE IC; AI1C. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/204650
28. Online Mendelian Inheritance in Man (OMIM). #301200 AMELOGENESIS IMPERFECTA, TYPE IE; AI1E. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/301200
29. Online Mendelian Inheritance in Man (OMIM). %301201 AMELOGENESIS IMPERFECTA, HYPOPLASTIC/HYPOMATURATION, X-LINKED 2. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/301201
30. Online Mendelian Inheritance in Man (OMIM). #616270 AMELOGENESIS IMPERFECTA, TYPE IF; AI1F. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/616270
31. Online Mendelian Inheritance in Man (OMIM). #204690 AMELOGENESIS IMPERFECTA, TYPE IG; AI1G. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/204690
32. Online Mendelian Inheritance in Man (OMIM). #616221 AMELOGENESIS IMPERFECTA, TYPE IH; AI1H. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/616221
33. Online Mendelian Inheritance in Man (OMIM). #617297 AMELOGENESIS IMPERFECTA, TYPE IJ; AI1J. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/617297
34. Online Mendelian Inheritance in Man (OMIM). #620104 AMELOGENESIS IMPERFECTA, TYPE IK; AI1K. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/620104
35. Online Mendelian Inheritance in Man (OMIM). #204700 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA1; AI2A1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/204700
36. Online Mendelian Inheritance in Man (OMIM). #612529 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA2; AI2A2. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/612529
37. Online Mendelian Inheritance in Man (OMIM). #613211 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA3; AI2A3. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/613211
38. Online Mendelian Inheritance in Man (OMIM). #614832 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA4; AI2A4. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/614832
39. Online Mendelian Inheritance in Man (OMIM). #615887 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA5; AI2A5. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/615887
40. Online Mendelian Inheritance in Man (OMIM). #617217 AMELOGENESIS IMPERFECTA, HYPOMATURATION TYPE, IIA6; AI2A6. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/617217
41. Online Mendelian Inheritance in Man (OMIM). #130900 AMELOGENESIS IMPERFECTA, TYPE IIIA; AI3A. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/130900
42. Online Mendelian Inheritance in Man (OMIM). #617607 AMELOGENESIS IMPERFECTA, TYPE IIIB; AI3B. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/617607
43. Online Mendelian Inheritance in Man (OMIM). #618386 AMELOGENESIS IMPERFECTA, TYPE IIIC; AI3C. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/618386
44. Online Mendelian Inheritance in Man (OMIM). #104510 AMELOGENESIS IMPERFECTA, TYPE IV; AI4. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/104510
45. Online Mendelian Inheritance in Man (OMIM). #190320 TRICHODENTOOSSEOUS SYNDROME; TDO. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/190320
46. Online Mendelian Inheritance in Man (OMIM). #217080 JALILI SYNDROME. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/217080
47. Online Mendelian Inheritance in Man (OMIM). #226750 KOHLSCHUTTER-TONZ SYNDROME; KTZS. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/226750
48. Online Mendelian Inheritance in Man (OMIM). #615905 DEVELOPMENTAL AND EPILEPTIC ENCEPHALOPATHY 25 WITH AMELOGENESIS IMPERFECTA; DEE25. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/615905
49. Online Mendelian Inheritance in Man (OMIM). #601216 DENTAL ANOMALIES AND SHORT STATURE; DASS. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/601216
50. Online Mendelian Inheritance in Man (OMIM). #234580 HEIMLER SYNDROME 1; HMLR1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/234580
51. Online Mendelian Inheritance in Man (OMIM). #616617 HEIMLER SYNDROME 2; HMLR2. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/616617
52. Online Mendelian Inheritance in Man (OMIM). #618363 SHORT STATURE, AMELOGENESIS IMPERFECTA, AND SKELETAL DYSPLASIA WITH SCOLIOSIS; SSASKS. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/618363
53. Online Mendelian Inheritance in Man (OMIM). #259775 RAINE SYNDROME; RNS. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/259775
54. Online Mendelian Inheritance in Man (OMIM). #612782 IMMUNODEFICIENCY 9; IMD9. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/612782
55. Online Mendelian Inheritance in Man (OMIM). #612783 IMMUNODEFICIENCY 10; IMD10. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/612783
56. Online Mendelian Inheritance in Man (OMIM). #226650 EPIDERMOLYSIS BULLOSA, JUNCTIONAL 1A, INTERMEDIATE; JEB1A. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/226650
57. Online Mendelian Inheritance in Man (OMIM). #620789 APLASIA CUTIS-ENAMEL DYSPLASIA SYNDROME; ACED. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/620789
58. Online Mendelian Inheritance in Man (OMIM). #617915 HYPOTONIA, ATAXIA, DEVELOPMENTAL DELAY, AND TOOTH ENAMEL DEFECT SYNDROME; HADDTS. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/617915
59. Orphanet. Enamel-renal syndrome (ORPHA:1031). Available from: https://www.orpha.net/en/disease/detail/1031
60. Orphanet. Jalili syndrome (ORPHA:1873). Available from: https://www.orpha.net/en/disease/detail/1873
61. Orphanet. Amelocerebrohypohidrotic syndrome (ORPHA:1946). Available from: https://www.orpha.net/en/disease/detail/1946
62. Orphanet. Brachyolmia-amelogenesis imperfecta syndrome (ORPHA:2899). Available from: https://www.orpha.net/en/disease/detail/2899
63. Orphanet. Deafness-enamel hypoplasia-nail defects syndrome (Heimler syndrome) (ORPHA:3220). Available from: https://www.orpha.net/en/disease/detail/3220
64. Orphanet. Osteosclerotic bone dysplasia (Raine syndrome) (ORPHA:1832). Available from: https://www.orpha.net/en/disease/detail/1832
65. Orphanet. Congenital scalp aplasia cutis-enamel hypoplasia-developmental delay-intellectual disability syndrome (ORPHA:697356). Available from: https://www.orpha.net/en/disease/detail/697356
66. Orphanet. Global developmental delay-dental enamel defects-ataxia syndrome (ORPHA:714399). Available from: https://www.orpha.net/en/disease/detail/714399
67. Aldred MJ, Savarirayan R, Crawford PJ. Amelogenesis imperfecta: a classification and catalogue for the 21st century. Oral Dis. 2003;9(1):19-23. doi:10.1034/j.1601-0825.2003.00843.x
68. Laouina S. Amelogenesis imperfecta or molar-incisor hypomineralisation: a narrative review and diagnostic framework for differentiating a genetic from an environmental enamel defect in children. Cureus. 2026;18(7):e112445. doi:10.7759/cureus.112445
69. Witkop CJ Jr. Amelogenesis imperfecta, dentinogenesis imperfecta and dentin dysplasia revisited: problems in classification. J Oral Pathol. 1988;17(9-10):547-53. doi:10.1111/j.1600-0714.1988.tb01332.x
70. Lagerström M, Dahl N, Nakahori Y, Nakagome Y, Bäckman B, Landegren U, et al. A deletion in the amelogenin gene (AMG) causes X-linked amelogenesis imperfecta (AIH1). Genomics. 1991;10(4):971-5. doi:10.1016/0888-7543(91)90187-j
71. Mårdh CK, Bäckman B, Holmgren G, Hu JC, Simmer JP, Forsman-Semb K. A nonsense mutation in the enamelin gene causes local hypoplastic autosomal dominant amelogenesis imperfecta (AIH2). Hum Mol Genet. 2002;11(9):1069-74. doi:10.1093/hmg/11.9.1069
72. Hart PS, Hart TC, Michalec MD, Ryu OH, Simmons D, Hong S, et al. Mutation in kallikrein 4 causes autosomal recessive hypomaturation amelogenesis imperfecta. J Med Genet. 2004;41(7):545-9. doi:10.1136/jmg.2003.017657
73. Kim JW, Simmer JP, Hart TC, Hart PS, Ramaswami MD, Bartlett JD, et al. MMP-20 mutation in autosomal recessive pigmented hypomaturation amelogenesis imperfecta. J Med Genet. 2005;42(3):271-5. doi:10.1136/jmg.2004.024505
74. El-Sayed W, Parry DA, Shore RC, Ahmed M, Jafri H, Rashid Y, et al. Mutations in the beta propeller WDR72 cause autosomal-recessive hypomaturation amelogenesis imperfecta. Am J Hum Genet. 2009;85(5):699-705. doi:10.1016/j.ajhg.2009.09.014
75. Parry DA, Brookes SJ, Logan CV, Poulter JA, El-Sayed W, Al-Bahlani S, et al. Mutations in C4orf26, encoding a peptide with in vitro hydroxyapatite crystal nucleation and growth activity, cause amelogenesis imperfecta. Am J Hum Genet. 2012;91(3):565-71. doi:10.1016/j.ajhg.2012.07.020
76. Parry DA, Poulter JA, Logan CV, Brookes SJ, Jafri H, Ferguson CH, et al. Identification of mutations in SLC24A4, encoding a potassium-dependent sodium/calcium exchanger, as a cause of amelogenesis imperfecta. Am J Hum Genet. 2013;92(2):307-12. doi:10.1016/j.ajhg.2013.01.003
77. Kim JW, Seymen F, Lee KE, Ko J, Yildirim M, Tuna EB, et al. LAMB3 mutations causing autosomal-dominant amelogenesis imperfecta. J Dent Res. 2013;92(10):899-904. doi:10.1177/0022034513502054
78. Poulter JA, Murillo G, Brookes SJ, Smith CE, Parry DA, Silva S, et al. Deletion of ameloblastin exon 6 is associated with amelogenesis imperfecta. Hum Mol Genet. 2014;23(20):5317-24. doi:10.1093/hmg/ddu247
79. Smith CE, Murillo G, Brookes SJ, Poulter JA, Silva S, Kirkham J, et al. Deletion of amelotin exons 3-6 is associated with amelogenesis imperfecta. Hum Mol Genet. 2016;25(16):3578-87. doi:10.1093/hmg/ddw203
80. Parry DA, Smith CE, El-Sayed W, Poulter JA, Shore RC, Logan CV, et al. Mutations in the pH-sensing G-protein-coupled receptor GPR68 cause amelogenesis imperfecta. Am J Hum Genet. 2016;99(4):984-90. doi:10.1016/j.ajhg.2016.08.020
81. Seymen F, Kim YJ, Lee YJ, Kang J, Kim TH, Choi H, et al. Recessive mutations in ACPT, encoding testicular acid phosphatase, cause hypoplastic amelogenesis imperfecta. Am J Hum Genet. 2016;99(5):1199-205. doi:10.1016/j.ajhg.2016.09.018
82. Kim JW, Zhang H, Seymen F, Koruyucu M, Hu Y, Kang J, et al. Mutations in RELT cause autosomal recessive amelogenesis imperfecta. Clin Genet. 2019;95(3):375-83. doi:10.1111/cge.13487
83. Smith CEL, Whitehouse LLE, Poulter JA, Wilkinson Hewitt L, Nadat F, Jackson BR, et al. A missense variant in specificity protein 6 (SP6) is associated with amelogenesis imperfecta. Hum Mol Genet. 2020;29(9):1417-25. doi:10.1093/hmg/ddaa041
84. Pousette Lundgren G, Dahllöf G. Advances in clinical diagnosis and management of amelogenesis imperfecta in children and adolescents. J Dent. 2024;147:105149. doi:10.1016/j.jdent.2024.105149
85. Wright JT, Hong SP, Simmons D, Daly B, Uebelhart D, Luder HU. DLX3 c.561_562delCT mutation causes attenuated phenotype of tricho-dento-osseous syndrome. Am J Med Genet A. 2008;146A(3):343-9. doi:10.1002/ajmg.a.32132
86. Online Mendelian Inheritance in Man (OMIM). *300391 AMELOGENIN; AMELX. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/300391
87. 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):61-8. doi:10.1002/humu.23348. Ferramenta: https://variantvalidator.org (consulta GRCh38, MANE Select, 29 out 2026)
88. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 331-2.
89. Wang SK, Zhang H, Lin HC, Wang YL, Lin SC, Seymen F, et al. AMELX mutations and genotype-phenotype correlation in X-linked amelogenesis imperfecta. Int J Mol Sci. 2024;25(11):6132. doi:10.3390/ijms25116132
90. Online Mendelian Inheritance in Man (OMIM). *606585 ENAMELIN; ENAM. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/606585
91. Hart TC, Hart PS, Gorry MC, Michalec MD, Ryu OH, Uygur C, et al. Novel ENAM mutation responsible for autosomal recessive amelogenesis imperfecta and localised enamel defects. J Med Genet. 2003;40(12):900-6. doi:10.1136/jmg.40.12.900
92. Ozdemir D, Hart PS, Firatli E, Aren G, Ryu OH, Hart TC. Phenotype of ENAM mutations is dosage-dependent. J Dent Res. 2005;84(11):1036-41. doi:10.1177/154405910508401113
93. Online Mendelian Inheritance in Man (OMIM). *601259 AMELOBLASTIN ENAMEL MATRIX PROTEIN; AMBN. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/601259
94. Online Mendelian Inheritance in Man (OMIM). *610912 AMELOTIN; AMTN. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/610912
95. Online Mendelian Inheritance in Man (OMIM). *604629 MATRIX METALLOPROTEINASE 20; MMP20. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/604629
96. Online Mendelian Inheritance in Man (OMIM). *603767 KALLIKREIN-RELATED PEPTIDASE 4; KLK4. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/603767
97. Online Mendelian Inheritance in Man (OMIM). *150310 LAMININ, BETA-3; LAMB3. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/150310
98. Online Mendelian Inheritance in Man (OMIM). *147558 INTEGRIN, BETA-6; ITGB6. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/147558
99. Yin H, Jang S, Kim H, Simmer JP, Hu JC, Kim JW. Novel ITGB6 mutations causing amelogenesis imperfecta. Genes (Basel). 2026;17(4):431. doi:10.3390/genes17040431
100. Online Mendelian Inheritance in Man (OMIM). *611927 SCAFFOLDING CK1-ANCHORING PROTEIN H; SACK1H. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/611927
101. Resende KK, Amorim LS, de Paula LM, Leite AF, Mazzeu JF, Yamaguti PM, et al. Pre-eruptive coronal resorptions as a clinical feature of FAM83H-related amelogenesis imperfecta: insights from two Brazilian families. Calcif Tissue Int. 2026;117:75. doi:10.1007/s00223-026-01540-8
102. Wright JT, Torain M, Long K, Seow K, Crawford P, Aldred MJ, et al. Amelogenesis imperfecta: genotype-phenotype studies in 71 families. Cells Tissues Organs. 2011;194(2-4):279-83. doi:10.1159/000324339
103. Online Mendelian Inheritance in Man (OMIM). *613214 WD REPEAT-CONTAINING PROTEIN 72; WDR72. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/613214
104. Online Mendelian Inheritance in Man (OMIM). *609840 SOLUTE CARRIER FAMILY 24 (SODIUM/POTASSIUM/CALCIUM EXCHANGER), MEMBER 4; SLC24A4. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/609840
105. Online Mendelian Inheritance in Man (OMIM). *601404 G PROTEIN-COUPLED RECEPTOR 68; GPR68. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/601404
106. Online Mendelian Inheritance in Man (OMIM). *614829 ODONTOGENESIS-ASSOCIATED PHOSPHOPROTEIN; ODAPH. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/614829
107. Online Mendelian Inheritance in Man (OMIM). *606362 ACID PHOSPHATASE 4; ACP4. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/606362
108. Liu L, Au CW, Hany U, Rigby AL, Chauhan A, Brown C, et al. ACP4 variants in hypoplastic amelogenesis imperfecta. Calcif Tissue Int. 2026;117:56. doi:10.1007/s00223-026-01512-y
109. Online Mendelian Inheritance in Man (OMIM). *611211 RECEPTOR EXPRESSED IN LYMPHOID TISSUES; RELT. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/611211
110. Nikolopoulos G, Smith CEL, Brookes SJ, El-Asrag ME, Brown CJ, Patel A, et al. New missense variants in RELT causing hypomineralised amelogenesis imperfecta. Clin Genet. 2020;97(5):688-95. doi:10.1111/cge.13721
111. Resende KKM, Riou MC, Yamaguti PM, Fournier B, Rondeau S, Pacot L, et al. Oro-dental phenotyping and report of three families with RELT-associated amelogenesis imperfecta. Eur J Hum Genet. 2023;31(11):1337-41. doi:10.1038/s41431-023-01440-7
112. Online Mendelian Inheritance in Man (OMIM). *608613 TRANSCRIPTION FACTOR Sp6; SP6. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/608613
113. Kim YJ, Lee Y, Zhang H, Song JS, Hu JC, Simmer JP, et al. A novel de novo SP6 mutation causes severe hypoplastic amelogenesis imperfecta. Genes (Basel). 2021;12(3):346. doi:10.3390/genes12030346
114. Online Mendelian Inheritance in Man (OMIM). *611062 FAMILY WITH SEQUENCE SIMILARITY 20, MEMBER A; FAM20A. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/611062
115. de la Dure-Molla M, Quentric M, Yamaguti PM, Acevedo AC, Mighell AJ, Vikkula M, et al. Pathognomonic oral profile of enamel renal syndrome (ERS) caused by recessive FAM20A mutations. Orphanet J Rare Dis. 2014;9:84. doi:10.1186/1750-1172-9-84
116. Eid MT, de Mul A, Muresan-Vintila L, Derain Dubourg L, Bertholet-Thomas A, Molin A, et al. Enamel renal syndrome due to FAM20A mutations: challenging kidney management in view of nephrocalcinosis, hypophosphatemia and hypocalciuria. Orphanet J Rare Dis. 2026;21:90. doi:10.1186/s13023-026-04232-6
117. Baker A, Al Tarrass M, Chatziantoniou C, Kozyraki R. FAM20C and FAM20A in normal and ectopic mineralization: a focus on oro-renal syndromes. Matrix Biol. 2025;142:58-85. doi:10.1016/j.matbio.2025.11.002
118. Online Mendelian Inheritance in Man (OMIM). *600525 DISTAL-LESS HOMEOBOX 3; DLX3. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/600525
119. Price JA, Bowden DW, Wright JT, Pettenati MJ, Hart TC. Identification of a mutation in DLX3 associated with tricho-dento-osseous (TDO) syndrome. Hum Mol Genet. 1998;7(3):563-9. doi:10.1093/hmg/7.3.563
120. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 770-1.
121. Nieminen P, Lukinmaa PL, Alapulli H, Methuen M, Suojärvi T, Kivirikko S, et al. DLX3 homeodomain mutations cause tricho-dento-osseous syndrome with novel phenotypes. Cells Tissues Organs. 2011;194(1):49-59. doi:10.1159/000322561
122. Li Y, Han D, Zhang H, Liu H, Wong S, Zhao N, et al. Morphological analyses and a novel de novo DLX3 mutation associated with tricho-dento-osseous syndrome in a Chinese family. Eur J Oral Sci. 2015;123(4):228-34. doi:10.1111/eos.12197
123. Bonnet AL, Sceosole K, Vanderzwalm A, Silve C, Collignon AM, Gaucher C. "Isolated" amelogenesis imperfecta associated with DLX3 mutation: a clinical case. Case Rep Genet. 2020;2020:8217919. doi:10.1155/2020/8217919
124. Whitehouse LLE, Smith CEL, Poulter JA, Brown CJ, Patel A, Lamb T, et al. Novel DLX3 variants in amelogenesis imperfecta with attenuated tricho-dento-osseous syndrome. Oral Dis. 2019;25(1):182-91. doi:10.1111/odi.12955
125. Wang YL, Lin HC, Chen JT, Chang HH, Hsieh TF, Hung HY, et al. Differential effects of DLX3 mutations drive phenotypic variability in tricho-dento-osseous syndrome via direct activation of WNT10A. Ann N Y Acad Sci. 2026;1557(1):e70228. doi:10.1111/nyas.70228
126. Duverger O, Ohara T, Bible PW, Zah A, Morasso MI. DLX3-dependent regulation of ion transporters and carbonic anhydrases is crucial for enamel mineralization. J Bone Miner Res. 2017;32(3):641-53. doi:10.1002/jbmr.3022
127. Haldeman RJ, Cooper LF, Hart TC, Phillips C, Boyd C, Lester GE, et al. Increased bone density associated with DLX3 mutation in the tricho-dento-osseous syndrome. Bone. 2004;35(4):988-97. doi:10.1016/j.bone.2004.06.003
128. Al-Batayneh OB. Tricho-dento-osseous syndrome: diagnosis and dental management. Int J Dent. 2012;2012:514692. doi:10.1155/2012/514692
129. Parry DA, Mighell AJ, El-Sayed W, Shore RC, Jalili IK, Dollfus H, et al. Mutations in CNNM4 cause Jalili syndrome, consisting of autosomal-recessive cone-rod dystrophy and amelogenesis imperfecta. Am J Hum Genet. 2009;84(2):266-73. doi:10.1016/j.ajhg.2009.01.009
130. Schossig A, Bloch-Zupan A, Lussi A, Wolf NI, Raskin S, Cohen M, et al. SLC13A5 is the second gene associated with Kohlschütter-Tönz syndrome. J Med Genet. 2017;54(1):54-62. doi:10.1136/jmedgenet-2016-103988
131. Ratbi I, Falkenberg KD, Sommen M, Al-Sheqaih N, Guaoua S, Vandeweyer G, et al. Heimler syndrome is caused by hypomorphic mutations in the peroxisome-biogenesis genes PEX1 and PEX6. Am J Hum Genet. 2015;97(4):535-45. doi:10.1016/j.ajhg.2015.08.011
132. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 151.
133. Lacruz RS, Feske S. Diseases caused by mutations in ORAI1 and STIM1. Ann N Y Acad Sci. 2015;1356(1):45-79. doi:10.1111/nyas.12938
134. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 577-8.
135. Online Mendelian Inheritance in Man (OMIM). #240300 AUTOIMMUNE POLYENDOCRINE SYNDROME, TYPE I, WITH OR WITHOUT REVERSIBLE METAPHYSEAL DYSPLASIA; APS1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/240300
136. Taylor AK, Lebwohl B, Snyder CL, Green PHR. Celiac disease. 2008 Jul 3 [updated 2025 Dec 4]. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1727/
137. Dourado MR, Dos Santos CRR, Dumitriu S, Iancu D, Albanyan S, Kleta R, et al. Enamel renal syndrome: a novel homozygous FAM20A founder mutation in 5 new Brazilian families. Eur J Med Genet. 2019;62(11):103561. doi:10.1016/j.ejmg.2018.10.013
138. Hassib NF, Shoeib MA, ElSadek HA, Wali ME, Mostafa MI, Abdel-Hamid MS. Two new families with enamel renal syndrome: a novel FAM20A gene mutation and review of literature. Eur J Med Genet. 2020;63(11):104045. doi:10.1016/j.ejmg.2020.104045
139. Goodspeed K, Liu JS, Nye KL, Prasad S, Sadhu C, Tavakkoli F, et al. SLC13A5 deficiency disorder: from genetics to gene therapy. Genes (Basel). 2022;13(9):1655. doi:10.3390/genes13091655
140. Ozlu C, Adams RM, Solidum RM, Cooper S, Best CR, Elacio J, et al. Developmental phenotype and quality of life in SLC13A5 citrate transporter disorder. Dev Med Child Neurol. 2025;67(7):930-40. doi:10.1111/dmcn.16218
141. Meng L, Huang D, Xie L, Song X, Luo H, Gui J, et al. Perampanel effectiveness in treating ROGDI-related Kohlschütter-Tönz syndrome: first reported case in China and literature review. BMC Med Genomics. 2023;16(1):292. doi:10.1186/s12920-023-01728-z
142. Bennett R, Rodd H, Monteiro J. An investigation of dental anomalies in individuals with amelogenesis imperfecta. Eur Arch Paediatr Dent. 2026 Oct 23. doi:10.1007/s40368-026-01302-y
143. Adorno-Farias D, Ortega-Pinto A, Gajardo P, Salazar A, Morales-Bozo I, Werlinger F, et al. Diversity of clinical, radiographic and genealogical findings in 41 families with amelogenesis imperfecta. J Appl Oral Sci. 2019;27:e20180359. doi:10.1590/1678-7757-2018-0359. PMID: 30970114. PMCID: PMC6442841.
144. Coffield KD, Phillips C, Brady M, Roberts MW, Strauss RP, Wright JT. The psychosocial impact of developmental dental defects in people with hereditary amelogenesis imperfecta. J Am Dent Assoc. 2005;136(5):620-30.
145. Altaher J, Parekh S, Ryan F. Evaluating a condition-specific PROM for amelogenesis imperfecta in children and young people: insights from patients and clinicians. Eur Arch Paediatr Dent. 2026;27:1425-33. doi:10.1007/s40368-026-01214-x
146. Farias MLM, Ornela GO, de Andrade RS, Martelli DRB, Dias VO, Júnior HM. Enamel renal syndrome: a systematic review. Indian J Nephrol. 2021;31(1):1-8. doi:10.4103/ijn.IJN_27_19
147. Koç NS. Stone-dominant renal phenotype without nephrocalcinosis in FAM20A-related enamel renal syndrome. Kidney Int Rep. 2026;11(4):103783. doi:10.1016/j.ekir.2026.103783
148. Roomaney IA, Kabbashi S, Chetty M. The craniofacial, dental and systemic manifestations of enamel renal syndrome: a scoping review. Eur J Med Genet. 2025;75:105008. doi:10.1016/j.ejmg.2025.105008
149. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 683-4.
150. Drumond VZ, de Arruda JAA, de Andrade BAB, Diniz-Freitas M, Mesquita RA, Abreu LG. Dento-craniofacial features of tricho-dento-osseous syndrome: a systematic review and meta-analysis. Spec Care Dentist. 2025;45(3):e70058. doi:10.1111/scd.70058
151. Islam M, Lurie AG, Reichenberger E. Clinical features of tricho-dento-osseous syndrome and presentation of three new cases: an addition to clinical heterogeneity. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;100(6):736-42. doi:10.1016/j.tripleo.2005.04.017
152. Nguyen T, Phillips C, Frazier-Bower S, Wright T. Craniofacial variations in the tricho-dento-osseous syndrome. Clin Genet. 2013;83(4):375-9. doi:10.1111/j.1399-0004.2012.01907.x
153. Kim YJ, Seymen F, Koruyucu M, Kasimoglu Y, Gencay K, Shin TJ, et al. Unexpected identification of a recurrent mutation in the DLX3 gene causing amelogenesis imperfecta. Oral Dis. 2016;22(4):297-302. doi:10.1111/odi.12439
154. Hirji N, Bradley PD, Li S, Vincent A, Pennesi ME, Thomas AS, et al. Jalili syndrome: cross-sectional and longitudinal features of seven patients with cone-rod dystrophy and amelogenesis imperfecta. Am J Ophthalmol. 2018;188:123-30. doi:10.1016/j.ajo.2018.01.029
155. Ozlu C, Spelbrink EM, Brown TL, Nye KL, Solidum RM, Cooper S, et al. SLC13A5 citrate transporter disorder epilepsy phenotype. Epilepsy Res. 2026;220:107718. doi:10.1016/j.eplepsyres.2025.107718
156. Liepina L, Kalnina ML, Micule I, Gailite L, Rots D, Kalnina J, et al. Kohlschütter-Tönz syndrome: case report with novel feature and detailed review of features associated with ROGDI variants. Am J Med Genet A. 2022;188(4):1263-79. doi:10.1002/ajmg.a.62613
157. Biçer BB, Aykan HH, Soğukpınar M, Şimşek Kiper PÖ. Aortic root dilatation and mitral valve prolapse in three siblings with dental anomalies and short stature syndrome due to a homozygous novel LTBP3 variant. Cardiol Young. 2025;35(5):1046-50. doi:10.1017/S1047951125001520
158. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 41, section 41.1 Glycosaminoglycans synthesis defects.
159. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 43, section 43.1 Congenital Disorders of Protein N-Glycosylation.
160. Schaaf C, Sailer S. USP7-related Hao-Fountain syndrome. 2025 Dec 4. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK619577/
161. Pousette Lundgren G, Vestlund GM, Dahllöf G. Long-term outcomes of ceramic crowns in adolescents and young adults with amelogenesis imperfecta: final follow-up of a randomized split-mouth controlled clinical trial. J Dent. 2026;173:106794. doi:10.1016/j.jdent.2026.106794
162. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 410.
163. Barak T, Kwan KY, Louvi A, Demirbilek V, Saygı S, Tüysüz B, et al. Recessive LAMC3 mutations cause malformations of occipital cortical development. Nat Genet. 2011;43(6):590–4. doi:10.1038/ng.836. PMID: 21572413.
164. Krämer S, Lucas J, Gamboa F, Peñarrocha Diago M, Peñarrocha Oltra D, Guzmán-Letelier M, et al. Clinical practice guidelines: oral health care for children and adults living with epidermolysis bullosa. Spec Care Dentist. 2020;40(S1):3–81. doi:10.1111/scd.12511.
165. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 827.
166. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, section 1.6.16 Stomatology.
167. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 679.
168. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 392-5.
169. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 448.
170. 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, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK572295/
171. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 535.
172. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 117.
173. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 568.
174. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 1022.
175. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 647.
176. American Academy of Pediatric Dentistry. Fluoride therapy. The Reference Manual of Pediatric Dentistry. Chicago (IL): American Academy of Pediatric Dentistry; 2025. p. 372-8. Available from: https://www.aapd.org/globalassets/media/policies_guidelines/bp_fluoridetherapy25.pdf
177. College ter Beoordeling van Geneesmiddelen (CBG-MEB). Samenvatting van de productkenmerken: Duraphat 50 mg/ml, suspensie voor dentaal gebruik (RVG 10942). Utrecht: CBG-MEB; revised 30 June 2025. Available from: https://www.geneesmiddeleninformatiebank.nl/smpc/h10942_smpc.pdf
178. Rhaiem M, Chalbi M, Bousaid S, Zouaoui W, Chemli MA. Dental treatment approaches of amelogenesis imperfecta in children and young adults: a systematic review of the literature. J Esthet Restor Dent. 2024;36(6):881-91. doi:10.1111/jerd.13191
179. Agencia Española de Medicamentos y Productos Sanitarios (AEMPS). Ficha técnica: Ozalin 2 mg/ml solución oral en envase unidosis (n.º 85069). Madrid: AEMPS; revised March 2026. Available from: https://cima.aemps.es/cima/dochtml/ft/85069/FT_85069.html
180. Agencia Española de Medicamentos y Productos Sanitarios (AEMPS). Ficha técnica: Kalinox 50%/50% gas comprimido medicinal en balas (n.º 67701). Madrid: AEMPS; revised June 2024. Available from: https://cima.aemps.es/cima/dochtml/ft/67701/FT_67701.html
181. European Medicines Agency. Sibnayal (potassium citrate/potassium hydrogen carbonate): EPAR — product information (summary of product characteristics). Amsterdam: EMA; first authorisation 2021 Apr 30, renewal 2026 Jan 8. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/sibnayal
182. ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine (US). Search by condition: amelogenesis imperfecta, enamel renal syndrome, Jalili syndrome, tricho-dento-osseous, Kohlschutter-Tonz (API v2). Accessed 29 Oct 2026. Available from: https://clinicaltrials.gov
183. Cho ES, Kim YJ, Seymen F, Koruyucu M, Hu JC, Simmer JP, et al. Antisense oligonucleotides-mediated rescue of the AMELX splicing defects. J Dent Sci. 2026;21(3):1480-5. doi:10.1016/j.jds.2025.07.3468
184. Prasov L, Ullah E, Turriff AE, Warner BM, Conley J, Mark PR, et al. Expanding the genotypic spectrum of Jalili syndrome: novel CNNM4 variants and uniparental isodisomy in a north American patient cohort. Am J Med Genet A. 2020;182(3):493-7. doi:10.1002/ajmg.a.61484
185. Bäckman B, Holmgren G. Amelogenesis imperfecta: a genetic study. Hum Hered. 1988;38(4):189-206. doi:10.1159/000153785
186. Gupta SK, Saxena P, Jain S, Jain D. Prevalence and distribution of selected developmental dental anomalies in an Indian population. J Oral Sci. 2011;53(2):231-8. doi:10.2334/josnusd.53.231