Dentinogenesis Imperfecta and Dentin Dysplasias

Isolated hereditary dentine defects: Shields type II and type III dentinogenesis imperfecta and type II dentine dysplasia (DSPP), type I dentine dysplasia (SMOC2, VPS4B, SSUH2), and their relationship with the dentinogenesis imperfecta of osteogenesis imperfecta and with the dentine defects of the rickets.

Literature and regulatory information updated: 29 October 2026.

0. Initial Reference Table

Since the Shields classification of 1973, hereditary dentine defects have been divided into three types of dentinogenesis imperfecta (DGI) and two types of dentine dysplasia (DD) [1, 2]. The numbers of the Online Mendelian Inheritance in Man (OMIM) catalogue and of the Orphanet portal identify each entity, and the relevant European Reference Network (ERN) is indicated. Type I DGI is the dental manifestation of osteogenesis imperfecta (OI); the remaining forms are disorders confined to the dentition. The table brings together the entities covered in this article.

1. Introduction

1.1. Definitions

Dentine is the mineralised tissue of ectomesenchymal origin that surrounds the pulp and supports the enamel; it consists of about 70% hydroxyapatite, 20% organic matrix and 10% water, and the organic matrix is made up of 85–90% type I collagen and, for the remainder, non-collagenous proteins, of which dentine phosphoprotein accounts for about half [2, 10]. Dentinogenesis imperfecta is a hereditary disorder of dentine formation and mineralisation that arises at the histodifferentiation stage of tooth development; the teeth acquire an opalescent colour ranging from blue-grey to yellow-brown, because the abnormal dentine shows through the enamel, and the unsupported enamel fractures and exposes dentine that wears away quickly [2, 24]. Dentine dysplasia is a group of hereditary dentine disorders in which the characteristic changes mainly affect the root, the pulp chamber and the root canals, with crowns of normal or near-normal appearance [22, 24]. In current practice, the two terms cover two distinct aetiological families: the disorders caused by DSPP variants, which form a continuum of severity, and radicular dentine dysplasia, which is genetically heterogeneous [10, 19].

1.2. First descriptions and the Brandywine isolate

Hereditary opalescent dentine has been described in large families since the early twentieth century; OMIM records the families published by Roberts and Schour in 1939, by Johnson and colleagues in 1959, the six-generation family of Bixler and colleagues in 1969, that of Giansanti and Budnick in 1975 and the five-generation family of Mars and colleagues in 1976 [3]. In 1957 Witkop estimated a frequency of 1 in 6,000 to 8,000 children in the United States, and in 1971 Witkop and Rao proposed reserving the term "dentinogenesis imperfecta" for the form associated with OI and calling the isolated form "opalescent dentine" [3, 16]. In 1956 Hursey, Witkop and colleagues described a form of DGI in the tri-racial Brandywine isolate in southern Maryland, a population in which, according to de La Dure-Molla and colleagues, genetic dental disorders as a whole reached an incidence of 1 in 15 [4, 10]. Radicular dentine dysplasia was first described as "rootless teeth" in a 1920 dissertation by Ballschmiede, and Rushton called it "dentinal dysplasia" in 1939 [7, 19].

1.3. The Shields classification and reclassifications

In 1973 Shields, Bixler and el-Kafrawy proposed the classification that still organises clinical practice: type I DGI, associated with OI; type II DGI, isolated hereditary opalescent dentine; type III DGI, the Brandywine form with shell teeth; type I DD, radicular; and type II DD, coronal [1, 2]. Witkop's nomenclature coexisted with it: Shields type I corresponds to Witkop's "dentinogenesis imperfecta", type II to "hereditary opalescent dentine" and type III to the "Brandywine isolate" [24]. OMIM removed DGI with OI from this series and renamed the former DGI-II as DGI1 (125490), keeping the numbering of DGI-III (125500), DD-I (125400) and DD-II (125420), which explains why "DGI1" and "DGI-II" coexist in the literature for the same disorder [2]. In 1975 Witkop suggested that types II and III were the same disorder, on the basis of a shared surname and clinical similarity, and in 1983 Levin and colleagues argued for keeping them separate, describing enamel pitting, enlarged pulps and anterior open bite in type III [4].

Genetics proved Witkop right. In 2001 two Chinese groups published in the same issue of Nature Genetics the first DSPP variants in families with DGI-II: Zhang and colleagues found a nonsense variant in exon 3, and Xiao and colleagues described three families, two of which also had progressive sensorineural hearing loss [17, 26]. In 2002 Rajpar and colleagues showed that DD-II also results from a DSPP variant [27], and in 2005 Kim and colleagues showed that the same variant, p.(Val18Phe), produced the type II phenotype in a Caucasian family and the type III phenotype in a Korean patient, concluding that types II and III are variations of a single disorder [28]. In 2015 de La Dure-Molla, Fournier and Berdal reviewed all published cases and proposed calling all DSPP disorders "dentinogenesis imperfecta", with three grades of severity (mild, the former DD-II; moderate, the former DGI-II; severe, the former DGI-III), excluding type I DGI because it belongs to OI and reserving "radicular dentine dysplasia" for the former DD-I [10]. The systematic review by Gilani and colleagues found that 37 of 41 family genetic studies still used the Shields classification, considered more practical in the clinic [29], and the 2026 Chinese consensus further proposed that DGI-III be regarded as an early phase of DGI-II, after following patients whose shell teeth progressed to complete pulp obliteration in adulthood [22].

1.4. Radicular dentine dysplasia: from an entity without a gene to a heterogeneous disorder

Until 2011 DD-I had no known cause, and in 2007 Kim and Simmer described it as the only form "untouched by the genetic revolution" [30]. That year Bloch-Zupan and colleagues identified homozygous SMOC2 variants in a consanguineous Turkish family with DD, microdontia and oligodontia, variants that had escaped exome sequencing and were found only by homozygosity mapping [31]. VPS4B followed, in a Chinese family with ten affected members (Yang and colleagues, 2016), and SSUH2, in a six-generation Chinese family (Xiong and colleagues, 2017) [32, 33]. There are also families with DD-I of recessive or dominant inheritance with no variant in any of these genes, which shows that the list is incomplete [12, 34].

2. Molecular Genetics and Pathophysiology

2.1. Dentinogenesis and the non-collagenous proteins of dentine

Odontoblasts differentiate at the bell stage of tooth development, form a single layer at the periphery of the pulp and secrete an unmineralised organic matrix, predentine, 10 to 40 µm thick, which is mineralised at a front that advances as the cells retreat towards the pulp [2]. The first dentine formed in the crown, mantle dentine, is 15 to 20 µm thick and rests on thick fibrils of type III collagen; it is followed by primary or circumpulpal dentine, which is denser, with thin fibrils of type I collagen; secondary dentine forms after root completion and tertiary dentine in response to caries or trauma [2]. Collagen provides the three-dimensional scaffold; dentine phosphoprotein (DPP, also called phosphophoryn), secreted by the odontoblast processes, binds to collagen fibrils in the transition zone between the overlap and gap segments and acts as a nucleator of hydroxyapatite [2, 10]. When immobilised on a support, DPP initiates hydroxyapatite formation at low concentration and inhibits crystal growth at high concentration, which gives it a role in regulating the size and shape of the crystallites [10]. In mutant mice, dentine sialoprotein (DSP) mainly regulates the initiation of mineralisation and DPP its maturation [29].

2.2. The DSPP gene and its products

The DSPP gene lies at 4q22.1 and has five exons spanning about 8.3 kb, the first of which is non-coding; exons 2 to 4 encode the signal peptide and most of DSP, and exon 5 encodes the carboxyl terminus of DSP, dentine glycoprotein (DGP) and all of DPP [2, 10, 11]. The MANE Select (Matched Annotation from NCBI and EMBL-EBI) reference transcript is NM_014208.3 (protein NP_055023.2) [35]. The primary human product has 1,253 amino acids and 124.6 kDa and is cleaved by proteases of the astacin and matrix metalloproteinase families into DSP (amino acids 16 to 374), DGP (375 to 462) and DPP (the remainder), present in mature dentine at a molar ratio of about 1 DSP to 10 DPP [2, 10, 11]. DSPP belongs to the SIBLING family (small integrin-binding ligand N-linked glycoprotein), five genes for secreted phosphoproteins clustered within about 375 kb on chromosome 4 (SPP1, DMP1, IBSP, MEPE and DSPP), phosphorylated by the Golgi kinase FAM20C at S-x-E motifs; DSPP has about 200 phosphoserines [11]. Exon 5 contains more than 200 tandem copies of a nominal 9 bp repeat encoding aspartic acid and phosphoserine sequences (DSS domain), which are highly negatively charged and bind calcium [20]. DSPP is expressed mainly in differentiated odontoblasts, transiently in presecretory ameloblasts and at levels about 400 times lower in bone, periodontal ligament, the mouse inner ear, kidney and eccrine sweat glands; none of these sites of expression has a known clinical phenotype outside the dentition [10, 11]. The absence of skeletal changes in patients with DSPP variants has been attributed to low expression in bone, to differences in processing or to redundancy of other matrix proteins [2].

2.3. Mechanisms: two classes of variants at the ends of the gene

Pathogenic DSPP variants cluster at two sites with distinct mechanisms [15, 23]. The first class affects the amino terminus: missense variants in the signal peptide (15 amino acids, MKIITYFCIWAVAWA) or in the IPV motif that follows it, nonsense variants in exon 3, and splicing variants in introns 2 and 3 that cause skipping of exon 3 [23, 36]. Rajpar and colleagues showed that a substitution in the signal peptide prevents translocation of the protein into the endoplasmic reticulum (ER) [27], and von Marschall and colleagues demonstrated, in HEK293 cells, that p.(Pro17Thr) and p.(Val18Phe) disrupt the IPV motif and retain the protein in the ER, that p.(Ala15Val) prevents removal of the leader peptide, which then anchors the protein to the ER membrane, and that all three mutants disturb the trafficking and processing of normal DSPP when co-expressed — a dominant-negative effect [11]. The second class lies in exon 5: deletions that shift the reading frame by −1 and convert the hydrophilic Ser-Ser-Asp repeats into a long hydrophobic sequence rich in valine, alanine and isoleucine, ending at a downstream stop codon [2, 37]. The mutant protein forms calcium-dependent aggregates in the rough ER that sequester normal DSPP, which constitutes a toxic gain of property with a dominant-negative effect [23]. In a mouse model carrying one of these variants, the odontoblasts lose polarity, activate autophagy, have abnormal junctions and lack processes, dentine deposition is delayed and ectopic mineralisation appears along the collagen [20].

It follows that the mechanism is not haploinsufficiency: the dosage curation of the ClinGen (Clinical Genome Resource) consortium assigns DSPP a haploinsufficiency score of 0, the mouse heterozygous for a null allele has a normal dentition, and only the homozygous null mouse reproduces the human disorder [2, 15]. A whole-gene deletion is therefore not equivalent to a pathogenic exon 5 frameshift, and a classical loss-of-function variant should not be classified by analogy [15]. The Dspp-null mouse has enlarged pulp chambers, widened predentine, hypomineralisation, pulp exposure, an irregular mineralisation front and failure of calcospherite coalescence, with accumulation of biglycan and decorin in the predentine — a phenotype similar to that of DGI-III [10, 11]. Knock-in DsppP19L mice, equivalent to the human variant p.(Pro17Leu), accumulate the protein heavily in the ER, whereas mice with a −1 frameshift in exon 5 accumulate it less, in parallel with the greater clinical severity of the signal peptide variants [23].

2.4. Spectrum of DSPP variants

In 2008 Barron and colleagues listed about twenty variants; de La Dure-Molla and colleagues counted 38 in 2015 (28 with DGI-II, 8 with DD-II and 2 with DGI-III, none in the DGP region); Wang and colleagues cite about 86 in the HGMD (Human Gene Mutation Database) in 2025; and the systematic review by Boonyakanog and colleagues, which covered publications up to August 2025, analysed 70 distinct variants in 48 publications and 99 genotype–phenotype records [2, 10, 23, 36]. Exon 5 harbours 61% of the variants (43 of 70), almost all of them (42 of 43) frameshifts; introns 2 and 3 contain only splicing variants; exon 2 contains only variants of the signal peptide or its cleavage site; exon 3 has 57% splicing and 43% missense variants; and exon 4 (4% of the total) has mainly nonsense variants [23]. By type, frameshifts account for 60%, splicing variants for 22%, and missense, nonsense and signal peptide variants for 4 to 6% each [23].

The table lists the variants of historical or clinical relevance, normalised in VariantValidator on the MANE transcript NM_014208.3 and, where available, with their ClinVar classification [35, 38].

2.5. Genotype–phenotype correlations

As early as 2008, Barron and colleagues noted a broad correlation based on few cases: the more upstream deletions in exon 5, which produce the longest hydrophobic sequences, were associated with DD-II and the more downstream ones with DGI-II or III [2]. The review by Boonyakanog and colleagues quantified the pattern: DD-II was significantly associated with exon 5 variants and with frameshifts; DGI-III was associated with variants in exons 2 and 3, mainly missense, and has never been described with exon 5 variants; and DGI-II showed no association with any region or variant type, occupying an intermediate position [23]. A clinically normal permanent dentition occurred in 95.45% of DD-II records and in 49.06% of carriers of exon 5 variants; thistle-tube pulp chambers in 72.73% and pulp stones in 27.27% of DD-II; severe attrition in 84.62%, shell teeth in 30.77% and pulp exposure in 38.46% of DGI-III; complete pulp obliteration in 74.47% of DGI-II; and a need for prosthodontic treatment in 38.46% of DGI-III [23]. Dental sensitivity was rare (three patients), all with exon 4 variants, which were also associated with enlarged pulp cavities and thin roots [23]. Wang and colleagues cite an observation by McKnight and colleagues that longer altered sequences would give DD and shorter ones, below 500 codons, DGI-II [36]. The authors of the review propose a hybrid framework in which the clinical classification is retained for immediate decisions and the genotype adds prognostic information: an exon 5 frameshift, especially in the amino-terminal portion of DPP, predicts DD-II with a near-normal permanent dentition; a missense or splicing variant of the signal peptide or of the IPV motif warns of a severe phenotype; and DGI-II, which is genetically heterogeneous, requires full surveillance whatever the variant [23]. Age at diagnosis confounds the classification: shell teeth are most evident in the primary dentition and progress to obliteration in the permanent teeth, so a sporadic adult tends to be classified as DGI-II regardless of genotype [22, 23].

2.6. Penetrance, expressivity, polymorphisms and de novo variants

Penetrance of the DSPP disorders is practically complete: Barron and colleagues write that normal teeth are never found in DGI-II, and in the French cohort of Caravello and colleagues all but one of the index cases with DGI or DD had an identified DSPP variant [2, 20]. Expressivity varies within the same family and within the same individual: the same variant produced DGI-II and DGI-III phenotypes, and a Chinese family with DD-II had a 5-year-old child with amber discoloration and severe attrition, a mother with normal-looking permanent teeth and thistle-tube chambers, and a grandfather similar to his granddaughter [5, 20, 28]. The frequency of de novo variants is low; Kim and colleagues described a de novo variant in a Korean family [28]. In 1972 Shokeir described a probable homozygote with severity similar to that of some heterozygotes [3].

The DPP domain is hypervariable in the population: in 188 normal chromosomes McKnight and colleagues found 20 insertions or deletions due to polymerase slippage and 37 single nucleotide polymorphisms, mainly C>T transitions at CpG and CpNpG sites, and 37 haplotypes; Song and colleagues described 14 in-frame insertions or deletions, 6 missense and 5 synonymous polymorphisms in 110 controls, organised into 15 haplotypes [37, 41]. These in-frame length variations and isolated missense changes in DPP have no demonstrated pathogenic effect [36, 41]. The literature contains errors arising from this architecture: a variant described as causal in a Swedish family, c.202A>T p.(Arg68Trp), is a common polymorphism; a Brandywine family was attributed to DPP insertions and deletions that were polymorphisms, the actual cause being p.(Pro17Ser); and an insertion seen in eight French families by short reads was, with long reads and the T2T reference, the pathogenic deletion c.2652del [20, 23, 40]. The diversity of DPP haplotypes between patients and the reference genome contributes to these alignment errors [20].

2.7. Sensorineural hearing loss (DFNA39)

Xiao and colleagues described two Chinese families with DGI and bilateral progressive high-frequency sensorineural hearing loss, beginning between 20 and 30 years of age, with tinnitus in several affected members and balance disturbance in one, carrying p.(Pro17Thr) and p.(Val18Phe), and a third family without hearing loss with a variant at the splice donor of intron 3; they demonstrated Dspp transcripts in the mouse inner ear [6, 17]. The association remains weak: the same p.(Val18Phe) was not accompanied by apparent hearing loss in the families of Kim and colleagues or in the three French families of Caravello and colleagues; the systematic review found hearing loss with four variants (c.49C>A, c.52-3C>G, c.52G>T and c.3676del) with no statistical association with diagnosis, region or variant type; GenCC (Gene Curation Coalition) submitters disagree (definitive versus limited), and ClinGen validates only the relationship with DGI [15, 20, 23, 28]. Kim and Simmer allowed that the hearing loss might be secondary to deep overbite caused by attrition, a hypothesis that Barron and colleagues consider unresolved [2].

2.8. Radicular dentine dysplasia: SMOC2, VPS4B, SSUH2 and cases without a gene

SMOC2 (6q27; MANE transcript NM_001166412.2) encodes a secreted matricellular protein of the SPARC family, with calcium-binding domains, which promotes matrix assembly and angiogenesis [19, 35]. The founder family, Turkish and highly consanguineous, had two cousins (a 10-year-old boy and a 5-year-old girl) homozygous for c.84+1G>T, a splice donor variant of intron 1, with extreme microdontia, oligodontia (14 and 13 permanent teeth missing), extra cusps, globular crowns, double teeth, taurodontism, thin enamel and very short roots; the unaffected parents and siblings were heterozygous and the variant was absent from 112 controls [7, 31, 42]. The second family, Pakistani and consanguineous, had three siblings homozygous for c.681T>A p.(Cys227*) on transcript NM_022138.3, which corresponds to c.648T>A p.(Cys216*) on the MANE transcript, with 10 to 15 permanent teeth missing, microdontia, taurodontism and single-rooted molars with thistle-shaped pulps [35, 42, 43]. In these families inheritance is AR, and knockdown of smoc2 in zebrafish altered the pharyngeal teeth and the expression of dlx2, bmp2 and pitx2 [31]. In 2023 Putrino and colleagues described an Italian family in which the father and two children with typical DD-I, without microdontia, carried the same c.84+1G>T, with a pedigree interpreted as dominant and possible involvement of the paternal grandfather and aunt; the paper does not state the zygosity, and the finding contradicts the recessive inheritance of the founder form, so both patterns are recorded with the source for each [7, 44].

VPS4B (18q21.33) encodes an AAA-family ATPase that disassembles ESCRT-III polymers in the formation of multivesicular bodies and in other membrane scission events [18, 19]. Yang and colleagues mapped DD-I in a four-generation Chinese family with ten affected members to a 9.63 Mb region at 18q21.2-q21.33 (LOD 3.9) and found the intronic variant NM_004869.4:c.790+46C>G (published designation IVS7+46C>G), which creates a new splice donor, inserts 45 bp into the transcript and 15 amino acids into the ATP-binding domain, reduces mRNA and protein levels and alters cellular localisation; knockdown in zebrafish reduced the size and number of teeth, with partial rescue by human mRNA, and the protein appeared to act upstream of the Wnt/β-catenin pathway [8, 32]. The Vps4b heterozygous mouse does not reproduce the disorder (only the predentine is slightly widened) and the homozygote is embryonic lethal; no second family has been published, and OMIM marks the gene–phenotype relationship as provisional [18, 45]. The variant is listed in ClinVar as pathogenic for DD type IB [38].

SSUH2 (3p25.3, previously described at 3p26.1) encodes a nuclear protein of about 39 kDa, possibly a chaperone or transcriptional regulator, highly expressed in dental pulp stem cells [9, 19]. Xiong and colleagues linked DD-I in a six-generation Chinese family with 14 affected members to a region at 3p26.1-p24.3 and identified the heterozygous variant c.353C>A p.(Pro118Gln) on transcript NM_015931.2, which on the MANE transcript NM_001256748.3 corresponds to c.419C>A p.(Pro140Gln); the mutant protein had an altered structure and reduced levels, knock-in mice had partial pulp obliteration and altered expression of Dspp, Dmp1, Runx2, Pax9, Bmp2 and Dlx2, and knockdown in zebrafish produced absence of teeth, rescued by human mRNA [33, 35]. In 2026 Deng and colleagues described a second Chinese family, with four affected members, carrying the same variant [46].

A consanguineous Moroccan family with four affected members of both sexes, born to healthy parents, had typical DD-I of recessive inheritance, without linkage to the DSPP locus and without SMOC2 variants [12], and family genetic analysis of an 11-year-old girl with DD-I found no variants in any of the three genes [34]. The inheritance of DD-I is therefore no longer uniformly dominant: it is AD with VPS4B and SSUH2 and in most classical families, AR with SMOC2 in the form with microdontia and in the Moroccan family, and the gene remains unknown in some families [19, 47]. The older pathophysiological hypotheses — premature invagination of Hertwig's epithelial root sheath with ectopic dentine formation, or a defect in the differentiation of radicular odontoblasts — have not been resolved by the genes identified [10, 19].

2.9. Type I collagen and other causes of dysplastic dentine

Type I DGI results from COL1A1 or COL1A2 variants in OI and is associated mainly with qualitative variants, in particular glycine substitutions in the triple helix [48, 49]. In the Swedish cohort of Andersson and colleagues, 70% (7 of 10) of children with a glycine substitution in COL1A1 located downstream of p.Gly305 had DGI in both dentitions, compared with 0 of 7 with upstream variants, and the same occurred in COL1A2 downstream of p.Gly211 (8 of 10 versus 0 of 5) [49]. In the Korean cohort of Lee and colleagues, DGI occurred in 71.1% of patients with a glycine substitution and in 33.3% of those with other variants (odds ratio 4.92; 95% confidence interval [CI] 2.19–11.06), with exceptions to the amino-terminal threshold (p.Gly194Arg and p.Gly257Arg in COL1A1 with DGI) [50]. The systematic review by Gilani and colleagues found four studies, mostly of Asian families, with type I collagen glycine variants in patients classified as non-syndromic DGI, including p.(Gly337Ser) in two Thai individuals and p.(Gly391Ser) in nine individuals from Thailand and South Korea, which justifies including COL1A1 and COL1A2 in panels for patients with apparently isolated DGI [29]. Kim and Simmer stress that defects in the genes encoding the enzymes of post-translational modification and assembly of collagen may include DGI, and that the genes for the less abundant non-collagenous proteins (DMP1, IBSP, MEPE, SPP1) have not proved to cause isolated dentine disease [30]. Odontochondrodysplasia, caused by compound heterozygosity for TRIP11 (allelic with achondrogenesis type IA), combines DGI with spondylometaphyseal dysplasia, and its form 2, due to MIA3 (TANGO1), adds hearing loss and diabetes [51, 52].

2.10. Table of the types in the class

3. Clinical Spectrum and Manifestations by System

3.1. Features common to dentinogenesis imperfecta

In all types of DGI the teeth show variable, opalescent discoloration, from blue-grey to yellow-brown or amber, as the abnormal dentine shows through the enamel; the severity of discoloration and enamel fracture varies widely, even within a family [24]. The dentine–enamel junction is abnormal and the enamel fractures along the striae of Retzius, flaking off during chewing; the exposed, hypomineralised dentine wears quickly, from a few facets to complete loss of the crown, and without treatment the whole dentition may wear down to the gingiva [10, 24]. Exposed dentine takes on a glassy, hard appearance through sclerosis, and patients rarely complain of sensitivity [2]. Radiographically the crowns are bulbous, with marked cervical constriction, the roots are short and thin, and the pulp chambers and canals are partially or completely obliterated; the radiodensity of enamel and dentine is normal, unlike in hypocalcified amelogenesis imperfecta (AI) [2, 24]. Histology shows sparse, wide, irregularly oriented dentinal tubules, atubular areas, reduced microhardness and changes in phosphorus, magnesium, sodium and calcium on X-ray spectroscopy [10, 29].

The consequences accumulate over time: loss of occlusal vertical dimension, fracture, pulp exposure and infection, aesthetic impact and, in severe cases, early tooth loss, with functional, nutritional and psychosocial effects [15]. Unrestored wear may interfere with facial growth and with the temporomandibular joint [53]. The relationship with caries is controversial: the Oxford Desk Reference and the review by Barron and colleagues state that teeth with DGI do not appear to be more susceptible to caries, although Barron adds that exposed dentine is more vulnerable than enamel; the GeneReviews chapter on OI describes DGI as an independent risk factor for caries in OI, and the atlases of Spranger and of Smith mention greater susceptibility in OI [2, 48, 54–56]. De La Dure-Molla and colleagues describe frequent periodontal disease without caries and attribute it to permeability to bacteria, pulp necrosis due to obliteration and neutrophil recruitment by dentine proteins [10].

3.2. Type II dentinogenesis imperfecta (hereditary opalescent dentine)

In DGI-II both dentitions are equally affected, penetrance is practically complete and there is no bone fragility [2, 3]. The primary dentition is often more severely affected than the permanent one: in the 6-year-old child illustrated by de La Dure-Molla and colleagues, the primary teeth were opalescent brown, with severe attrition, complete obliteration, bulbous crowns, short roots and periapical radiolucency without caries, whereas the permanent teeth, at 8 years, were less affected [10]. Before and immediately after eruption, the pulp chambers are larger than normal and close progressively as abnormal dentine is deposited; obliteration may begin before eruption [15, 24]. In the three Brazilian families of Acevedo and colleagues, the colour changed with age (yellow to brown in primary teeth, blue-grey in young permanent teeth, brown in adult permanent teeth), six affected individuals had periapical lesions and the adults had extensive tooth loss; ground sections showed pulp stones in two families, suggesting that obliteration results from the growth of these structures [40]. In the French cohort of Caravello and colleagues, pulp infections were frequent and the roots short and conical [20].

3.3. Type III dentinogenesis imperfecta (Brandywine)

DGI-III was described in the Brandywine isolate; the crowns of primary and permanent teeth wear rapidly after eruption, the dentine is amber and smooth, and multiple pulp exposures may occur [4]. Radiographs of the primary teeth show, in the first years, very large chambers and canals, with thin dentine beneath the enamel and cementum — shell teeth — and the permanent teeth have smaller or obliterated pulp spaces [4, 22]. Levin and colleagues described enamel pitting, apical radiolucencies without pulp exposure and anterior open bite in all patients with a complete permanent dentition, and a localised form of juvenile periodontitis co-segregated with DGI in one family, at distinct but linked loci [3, 4]. The Chinese consensus followed a boy with shell teeth at 12 years in the canines and posterior teeth and obliteration in the earlier-erupted teeth, and at 29 years found complete obliteration of the whole dentition; in another family, the young patient had shell teeth and the adult obliteration [22]. De La Dure-Molla and colleagues counted enlarged pulps in five cases in the primary dentition (all between 2 and 4 years) and in one case in the permanent dentition, and recommend waiting for the permanent dentition for the definitive diagnosis [10]. In the 3-year-old proband of Wang and colleagues, wear led to multiple pulp exposures with pulpitis and apical periodontitis [36].

3.4. Type II dentine dysplasia (coronal)

DD-II reproduces in the primary dentition the picture of DGI-II (amber teeth, bulbous crowns, cervical constriction, thin roots, early obliteration), but the permanent teeth have normal colour, shape and height or only slight discoloration; radiographs of the permanent teeth show thistle-tube pulp chambers — a wide chamber extended by thin canals — and frequent pulp stones, without periapical radiolucencies, and the roots are of normal length [5, 10, 24]. The Chinese consensus also describes, in young permanent teeth, an X-shaped root canal, narrowed in the middle, which it interprets as the beginning of abnormal mineral deposition [22]. The normal-looking permanent dentition facilitates misdiagnosis when the patient is first seen as an adult, and DGI-II is often found in other family members or in the patient themselves [10, 22]. In the French series, four families had DD, with clinically normal-looking permanent teeth, wide chambers and, in one of them, flame-shaped chambers with calcifications [20].

3.5. DSPP-associated hearing loss

In families with DFNA39, bilateral sensorineural hearing loss began in the high frequencies between 20 and 30 years of age and progressed; several patients had tinnitus [6]. The Chinese consensus and Barron and colleagues mention hearing loss as a rare manifestation of DGI-II, and the evidence remains limited to a few families [2, 15, 22]. A single patient with fractures and osteoporosis and an exon 5 DSPP variant, with no collagen variant identified, has been described, without proof of a causal relationship with bone [23].

3.6. Type I dentine dysplasia (radicular)

In DD-I the crowns have normal or near-normal shape and colour (sometimes slightly opalescent, bluish, brown or with brown opacity of the incisal edge), and the first sign is mobility, with spontaneous exfoliation or exfoliation after minimal trauma; delayed eruption, discomfort due to mobility especially after meals, and pain from abscesses in caries-free teeth are described [10, 19]. Radiographs show short, conical, tapering roots with apical constriction, or absent roots; pre-eruptive pulp obliteration, complete in the primary teeth and with a crescent- or chevron-shaped remnant parallel to the cementoenamel junction in the permanent teeth; root canals that are generally invisible; and multiple periapical radiolucencies in caries-free teeth, corresponding to chronic abscesses, granulomas or cysts, which are the most discriminating sign [2, 7, 24]. The variation depends on when the radicular dentine becomes disorganised: early on, absent roots with no detectable pulp; later, shortened roots with a crescent-shaped chamber; late, roots of normal length with pulp stones in a normally shaped chamber; the variation is greater in the permanent dentition and from tooth to tooth [24]. O'Carroll and Duncan proposed four subtypes: DD1a, with no chamber or root and many periapical radiolucencies; DD1b, with horizontal crescent-shaped pulp remnants and a few millimetres of root; DD1c, with two horizontal crescent-shaped lines and a significant but incomplete root; and DD1d, with visible chambers and oval pulp stones in the coronal third of the canal, which bulge the root, and few radiolucencies [7, 19]. In the six-generation family studied by these authors, 35 people had confirmed or reliably reported DD-I and, of the 18 with a panoramic radiograph, 8 had DD1a and 10 DD1b; in the family of Kalk and colleagues, root length did not correlate with obliteration and subclassification was not possible [7]. The shorter the root, the greater the obliteration and periodontal pathology; the root may have a taurodont appearance [10]. Histology shows normal coronal dentine that, deeper down, becomes atubular, irregular and "sand dune"-like, filling the chamber [7, 19].

The age at tooth loss varies with the gene. In the VPS4B family, hypermobility began between 1.2 and 16 years, exfoliation between 16 and 20 years, and two patients were edentulous after 45 [8]. In the SSUH2 family, all adults had hypermobility from the age of 10 and lost their teeth a decade later, six of the oldest were edentulous at 35, the alveolar process was resorbed and gingival attachment was abnormal [9]. In the family of Kalk and colleagues, the proband's father had a complete upper denture and a partial lower denture by the end of adolescence [7]. Early exfoliation leads to maxillomandibular atrophy, which complicates rehabilitation [2, 19].

The SMOC2 form is distinguished by the extreme microdontia, oligodontia, shape anomalies and double teeth that accompany the radicular defect; the girl in the founder family was obese and had very mild skeletal changes, absent in her cousin, and psychomotor development was normal in both [7]. In the Pakistani family, two siblings had anterior open bite and all three had increased calculus [7]. The Chinese consensus adds that DD may be accompanied by bone sclerosis in other regions, which points to the syndrome of DD-I with sclerotic bones [19, 22].

3.7. Dentinogenesis imperfecta in osteogenesis imperfecta (type I DGI)

In OI, the primary dentition is the most affected, followed by the permanent incisors and first molars, with the second and third molars the least altered; radiographs show bulbous crowns, cervical constriction, thin roots, early obliteration, periapical radiolucencies and root fractures, and teeth with complete obliteration coexist with teeth of normal dentine in the same individual [2, 24]. According to GeneReviews, DGI is rare in classic non-deforming OI with blue sclerae (type I), constant in the perinatal lethal form and in the progressively deforming form (types II and III), and inconsistent in the common variable form with normal sclerae (type IV); Sillence designated the subgroup of type I with DGI as IB, although OMIM, citing Levin (1980), and the review by Barron use the letters the other way round [2, 48, 57]. The atlases agree on the pattern: the Oxford Desk Reference, reproducing the expanded Sillence classification, records absence of DGI in type I and in types V, VI and VII and DGI in types III and IV, and Spranger's atlas describes occasional DGI in type III, frequent DGI in type IV and great intra- and interfamilial variability in type I, of no use for subclassifying OI [55, 58–61]. The measured frequency depends on the cohort and the method:

The Romanian authors compare their figure with the 30–50% of Western European series and attribute the difference to ascertainment and method [63]. The presence or absence of DGI is almost always concordant between affected parents and children and between siblings [62]. In the Swedish cohort, DGI was confined to the primary dentition in 20 children (17 with subtle signs in the permanent teeth), taurodontism affected 18% and retention of the permanent second molars 31% (50% in qualitative defects versus 16% in quantitative ones) [49]. The recessive non-collagen forms of OI rarely have DGI (GeneReviews records absence of DGI in the forms due to BMP1, CRTAP, P3H1, PPIB, SERPINF1, SP7 and TMEM38B, and rare DGI in type V due to IFITM5) [48].

3.8. Dentine defects of the rickets

Vitamin D-dependent rickets and vitamin D-resistant rickets (now called hypophosphataemic rickets, mostly X-linked) produce dentine defects with large pulp chambers, extensive pulp horns, short roots, widened predentine and spontaneous abscesses in caries-free teeth, with chronic periodontal disease in the former and attrition and exposure of the dentine of the primary teeth in the latter [2, 19]. In X-linked hypophosphataemia (XLH), the German S3 guideline recommends six-monthly examinations from eruption of the first tooth to look for pulp inflammation or necrosis in teeth without caries or trauma [21]. These disorders include the hypophosphataemic rickets (X-linked and hereditary) and the vitamin D-dependent rickets.

4. Clinical Recognition and Warning Signs

4.1. When to suspect

Suspicion arises at eruption of the primary dentition when the teeth are opalescent, amber or blue-grey, with enamel that chips and rapid wear, especially if there is a family history with vertical transmission [2, 21]. The German S3 guideline lists as indicators of genetic dental disease an unusual and generalised constellation, primary teeth of atypical shape or structure, multisystem or syndromic disease, and similar findings in several relatives, and points out that in these disorders all teeth in both dentitions may look almost identical [21]. In DD-I, the child may present with tooth mobility, abscesses in caries-free teeth or delayed eruption, or as an incidental finding on a panoramic radiograph requested for orthodontics; in the family of Putrino and colleagues, the radiographic signs went unnoticed for years by several dentists [19, 44].

4.2. Signs requiring medical assessment

A child with opalescent teeth and fractures after minimal trauma, bluish sclerae, short stature, joint hypermobility or hearing loss should be assessed for OI, because DGI may be the most penetrant feature of the disorder and mild OI may have gone undiagnosed [2, 15]. The AAPD recommends looking for DGI in children with unexplained fractures, as an indicator of undiagnosed OI, to distinguish child abuse from mild OI [24], and Malmgren and Norgren showed that, in mild forms with an uncertain medical diagnosis, the dental changes may be decisive [62]. Idiopathic early loss of primary teeth requires assessment for hypophosphatasia [21]. Short roots and abscesses in caries-free teeth with limb bowing, short stature or signs of rickets require measurement of calcium, phosphate, alkaline phosphatase, parathyroid hormone and 25-hydroxyvitamin D before concluding that the DD is hereditary [19, 47]. DGI with short-limbed short stature, joint laxity and a narrow chest suggests odontochondrodysplasia [52].

4.3. Dental emergency signs

Pulp exposures and periapical abscesses, frequent in severe DGI and in DD-I, are the situations that require rapid intervention; in the DGI-III of the 3-year-old proband of Wang and colleagues, the multiple exposures required pulp treatment under general anaesthesia [24, 36]. There is no newborn screening for these disorders.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Clinical and radiographic diagnosis

Diagnosis rests on the history, clinical examination and radiography, and the panoramic radiograph is the central investigation [2, 19]. The history should confirm that the change is hereditary rather than acquired, construct the pedigree, characterise the primary dentition (colour, wear, abscesses, mobility, early loss) and look for signs of systemic disease: fractures, hyperextensibility, short stature, hearing loss, scleral colour and, in the context of DD-I, chemotherapy or irradiation of the jaws during root formation [2, 19]. Bulbous crowns, short thin roots and progressive pulp obliteration in opalescent teeth are almost pathognomonic of the phenotype but do not identify the gene or the subtype [15]. Because obliteration progresses, radiographic follow-up at several ages is recommended to establish the diagnosis, and de La Dure-Molla and colleagues advise waiting for the permanent dentition before classifying a young child with enlarged pulps as severe [2, 10]. Scanning electron microscopy and ground sections of exfoliated or extracted teeth are complementary; in the Swedish OI cohort, histology confirmed DGI in 54% of children without clinical or radiographic signs [49]. There is no clinical assay for dentine proteins [15].

The 2026 Chinese consensus formalises criteria for each type [22]:

The table by Barron and colleagues, built from Shields and from later series, adds that pulp exposures and shell teeth are described only in DGI-III, that periapical radiolucencies characterise DGI-III and DD-I, that short, conical and tapering roots or absent roots occur only in DD-I, that the thistle-tube chamber and pulp stones characterise the permanent dentition of DD-II, and that OI accompanies only DGI-I [2]. The German S3 guideline stresses that clinical diagnosis is visual and made by an experienced dentist, that radiographs taken on indication add information, and that only genetic analysis clarifies the aetiology, the prognosis and any involvement of other organs [21].

5.2. Additional medical investigations

In the presence of DGI, OI is excluded by the history of fractures, scleral colour, family history and, where relevant, densitometry and testing of COL1A1 and COL1A2 [15, 20]. In the presence of DD-I or short roots with abscesses in caries-free teeth, serum calcium, phosphate, alkaline phosphatase, parathyroid hormone and 25-hydroxyvitamin D and skeletal assessment exclude the rickets, and low alkaline phosphatase points to hypophosphatasia [21, 47]. Pure-tone audiometry is a reasonable precaution in carriers of DSPP variants in the signal peptide or at the start of exon 3, without constituting established surveillance [15].

5.3. Genetic tests and their limits

The German S3 guideline recommends genetic diagnosis when clinical and radiological findings suggest a genetic cause, usually in liaison with a genetics clinic (expert consensus; moderate recommendation), by massively parallel sequencing when a monogenic cause is suspected [21]. The main technical limitation is exon 5 of DSPP: the more than 200 repeats of 9 bp hamper Sanger and short-read sequencing, alignments produce artefacts and low coverage, and the hypervariability of DPP in the population generates insertions and deletions that may be mistaken for pathogenic variants [20, 37]. In the Strasbourg cohort (112 individuals, 42 index cases and 70 relatives, recruited between 2016 and 2024), the 653-gene GenoDENT panel, with additional probes for the DSPP repeats, found candidate variants in 31 of the 42 families, but only 7 could be validated directly; Sanger sequencing validated 8, long-range polymerase chain reaction (PCR) amplification of exon 5 followed by Sanger validated 6, and long-read nanopore sequencing (Oxford Nanopore Technology, ONT) validated 10; of the 11 panel-negative families, ONT resolved 10 [20]. Overall, a molecular diagnosis was reached in 41 families, with 22 pathogenic or likely pathogenic variants (8 known and 14 novel), and ONT contributed in 20 families [20]. The authors propose starting with long-range PCR of exon 5 with ONT sequencing and only then, if negative, moving on to a panel or exome [20]. A negative short-read panel or exome does not exclude DSPP in a patient with a convincing dentine phenotype [15].

Other limitations are common to all genes in the class. The SMOC2 founder family had a splicing variant that the exome did not detect [31]; deep intronic variants, regulatory variants, variants that are only apparent at the RNA level and variants in genes not included in panels may explain the undiagnosed cases [20, 23]; and the German guideline points out that a genetic result without abnormalities does not exclude a genetic aetiology [21]. A variant of uncertain significance (VUS) is not diagnostic and calls for family segregation studies or functional studies [20, 21]. Because the DSPP mechanism is not haploinsufficiency, a gene deletion or an isolated loss-of-function variant cannot be classified by analogy with exon 5 frameshifts [15]. The proportion of families with an identified variant in a selected cohort with access to long reads is not equivalent to the clinical sensitivity of a standard panel.

5.4. Content of the report

The report should identify the gene, the reference transcript with version (NM_014208.3 for DSPP; NM_001166412.2, NM_004869.4 and NM_001256748.3 for SMOC2, VPS4B and SSUH2), the variant in HGVS (Human Genome Variation Society) nomenclature, the zygosity, the class according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) and the technology that covered exon 5 [20, 35]. When the literature uses another numbering — for example, SSUH2 c.353C>A on transcript NM_015931.2, which is c.419C>A on the MANE transcript, or SMOC2 c.681T>A on NM_022138.3, which is c.648T>A on the MANE transcript — both descriptions should be recorded [35]. The classification of older variants should be reviewed: c.202A>T p.(Arg68Trp), still listed in OMIM as allelic, has a benign or likely benign classification in ClinVar [11, 38].

5.5. Algorithm

5.6. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, preimplantation genetic diagnosis after in vitro fertilisation may be considered in accordance with local ethical rules, according to the Chinese consensus [22]; GeneGuidance notes that confirmation of the gene allows prenatal and preimplantation options to be discussed [15]. The sources consulted do not describe actual use of these techniques in families with isolated dentine disorders.

6. Differential Diagnosis

6.1. Comparative table

6.2. Main differentials

The most important distinction, because of its medical consequences, is between isolated DGI and the DGI of OI: GeneGuidance stresses that a DSPP variant does not imply bone fragility and that a patient with fragile teeth and fractures needs testing of the collagen genes, not of DSPP [15]. DGI may be the most visible feature of mild OI, and Caravello and colleagues recommend genetic testing in all patients with dentine defects because of the possibility of undiagnosed OI or another skeletal disorder; the review by Gilani and colleagues shows that collagen glycine variants may present as DGI without other signs of OI [20, 29]. The Brazilian family described by Pallos and colleagues, initially diagnosed with DGI, had a COL1A1 variant and ligamentous hyperlaxity, and was reclassified as mild OI [40].

Among the DSPP forms, DD-II is distinguished from DGI-II by normally coloured permanent teeth, by chambers that do not obliterate before eruption and by the thistle-tube chamber in a single-rooted tooth; the crowns in DD are of normal size and shape, unlike the bell-shaped crowns with cervical constriction of DGI, and the roots in DD are absent (DD-I) or normal (DD-II), whereas those in DGI are short and narrow [24]. DD-I is distinguished from all forms of DGI by the normal crowns, and DGI-III shares with it the multiple periapical radiolucencies in caries-free teeth [2]. Early loss of teeth with short roots distinguishes DD-I from hypophosphatasia, in which the teeth are shed with complete roots [10]. Hypocalcified AI is distinguished by enamel that is less radiopaque than dentine and by dental sensitivity [2]. The syndromes with DGI mentioned by Barron and colleagues include Ehlers-Danlos syndrome, Schimke immuno-osseous dysplasia, brachio-skeleto-genital syndrome and a form of primordial osteodysplastic dwarfism with severe microdontia, opalescent teeth and rootless molars [2]; Orphanet also groups the syndrome of DGI with short stature, hearing loss and intellectual disability and the skeletal dysplasia with Wormian bones, multiple fractures and DGI [13]. Two 2026 reviews in the Journal of the American Dental Association brought together more than 60 genes and conditions in which dentine defects are accompanied by skeletal, ocular, auditory or renal comorbidities, and the dental manifestations of rare skeletal disorders (OI, XLH, hypophosphatasia, cleidocranial dysplasia), stressing that clinical overlap hampers diagnosis and requires multidisciplinary care [65, 66]. Bruck syndrome and spondylo-ocular dysplasia, differentials of OI, do not have DGI [67, 68].

7. Treatments and Risk Management Strategies

7.1. Goals and evidence base

There is no targeted pharmacological or gene treatment for hereditary dentine disorders; management is dental, preventive and restorative, and starts early because the damage is mechanical and cumulative [15, 21]. The goals are to eliminate sources of pain and infection, protect the posterior teeth from wear and maintain the occlusal vertical dimension, restore aesthetics and preserve the natural dentition for as long as possible, adapting the plan to age, severity and the presenting complaint [2, 22, 24]. The evidence is weak: the German S3 guideline, which searched the literature from 2000 to 2022, found only case reports and case series for DGI and graded all specific recommendations as very low evidence under the GRADE system (Grading of Recommendations, Assessment, Development and Evaluation), although the panel assigned them, by consensus, a strong strength [21]. The AAPD (American Academy of Pediatric Dentistry) based its 2013 guideline on 39 articles and on expert opinion where data were insufficient, and that guideline was withdrawn in 2018, remaining as a historical reference [24].

7.2. Recommendations of the German S3 guideline for DGI

The German S3 guideline on rare dental diseases, developed between 2021 and 2024 and published in German in early 2025 and as an English summary in 2026, contains 41 recommendations, 33 of them evidence-based and 8 consensus-based [21].

The guideline text adds that direct and indirect restorations reliably compensate for the defects, that definitive restoration of the whole permanent dentition is usually carried out at the end of the second or the beginning of the third decade, that the priority is to secure posterior function and vertical dimension before restoring the anterior teeth, and that rehabilitation, begun in childhood, ends only after growth is complete, with restorative needs throughout life [21]. From adolescence onwards, indirect restorations on the first permanent molars prevent loss of tooth substance, control pain and maintain the vertical dimension [21].

7.3. Prevention

Prevention is common to all sources: meticulous oral hygiene, calculus removal, mouthwashes, dietary advice to reduce sugar, professional cleaning, application of fluoride and desensitising agents, and periodic examinations that treat each tooth as it erupts [2, 21, 24]. The Chinese consensus recommends fissure sealants soon after emergence of the permanent teeth and visits every 3 to 6 months, more often in severe cases, because the risk of caries and periodontal disease persists after treatment [22]. In DD-I, early loss due to periodontitis is frequent because of the short roots, so hygiene must be established and rigorously maintained; in the family of Putrino and colleagues, visits were scheduled every 3 or 4 months after the diagnosis [24, 44]. Early diagnosis and regular care do not, however, prevent the tooth loss caused by short or absent roots and by spontaneous abscesses in some forms of DGI and DD [2].

7.4. Primary and mixed dentitions

In primary teeth with normal morphology and little wear, the Chinese consensus recommends strip crowns on the anterior teeth and stainless steel crowns on the posterior teeth as a preventive measure; with attrition, preparation should be minimal or none; with severe attrition, removable overdentures are the preferred option [22]. Barron and colleagues describe the same scheme — stainless steel crowns on the molars to prevent wear and maintain the vertical dimension, veneers or composite crowns for aesthetics, and overdentures when the child presents late with teeth worn to the gingiva — adding that children adapt well to overdentures, which need to be reviewed and remade with growth, and that pulp therapy of abscessed primary teeth is unsuccessful, so these should be extracted [2]. The review by Zhang and colleagues cites aesthetic preformed crowns (composite strip, polycarbonate or zirconia) on primary anterior teeth, with no evidence favouring any material [53]. Primary teeth with severe periapical infection should be extracted to protect the permanent tooth germs, with placement of a space maintainer adjusted to eruption [22]. The AAPD accepted that primary teeth could receive composite or veneered anterior crowns and steel or veneered posterior crowns [24].

In the mixed dentition, the goals are to preserve tooth structure, relieve sensitivity, maintain the vertical dimension and restore aesthetics; before reaching functional occlusion, the permanent teeth receive provisional occlusal onlays on the posterior teeth and composite on the anterior teeth, and full-coverage indirect restorations once occlusion is established and the cementoenamel junction is accessible [22]. General anaesthesia is an option when multiple teeth need treatment and the child cannot cooperate [2, 21, 22]. In the 3-year-old proband of Wang and colleagues, pulpotomy, pulpectomy and stainless steel or composite strip crowns were performed under general anaesthesia [36].

7.5. Adolescence and the transition phase

Between 12 and 18 years, with all permanent teeth erupted but roots still incomplete, the Chinese consensus maintains the provisional strategies of the mixed dentition, recommends removable prostheses when teeth with a poor prognosis must be extracted and reserves fixed prostheses for adulthood [22]. Barron and colleagues advise monitoring the rate of wear of the permanent teeth, intervening only when necessary, placing cast occlusal onlays on the first molars and then on the premolars, and keeping preparation minimal until adulthood [2]. For the shell teeth of DGI-III, the Chinese consensus recommends immediate provisional restorations, posterior onlays and anterior composite after eruption, and definitive restoration only after pulp obliteration, which appears with age [22].

7.6. Adult rehabilitation

In adults, the German S3 guideline and the Chinese consensus recommend crowns in DGI and complete occlusal reconstruction when the vertical dimension has been altered by severe attrition, and tooth-supported or implant-supported rehabilitation after tooth loss [21, 22]. Conventional restorative techniques are suitable in mild to moderate DGI; in cases with enamel fracture and rapid wear, full-coverage restoration is the treatment of choice, with greater success in teeth whose crown and root are close to normal, because short thin roots and marked cervical constriction increase the risk of cervical fracture [2, 24]. Ideally, stabilisation should be completed before wear and loss of vertical dimension become excessive; when crown loss is severe, overdentures over teeth covered with fluoride-releasing glass ionomer are an alternative [24]. Bleaching partially lightens the teeth but does not normalise the colour, which comes from the dentine; veneers mask anterior discoloration [24].

The review by Zhang and colleagues (92 articles, 28 tabulated cases) summarises the solutions described: composite and veneers when there is no enamel fracture; full crowns for extensive wear; self-threading parapulpal pins or crown lengthening (only in teeth without bulbous shape or short roots) to obtain retention; splinted crowns to reduce the risk of root fracture; indirect composite onlays with no adhesive failure at 13 years in one case; fixed bridges with an increased risk of root fracture; implants that are predictable but expensive; and overdentures that are cheaper but hamper hygiene [53]. The same authors treated 25-year-old twin sisters with DGI-II, complete obliteration and a poorly tolerated removable partial denture, with metal-ceramic crowns on posts and pin-lays prepared with printed guides planned by cone-beam computed tomography (CBCT) and computer-aided design and manufacturing (CAD/CAM), and two implants in one of them; there were decementations at 20, 23 and 37 months and one fracture at 40 months while chewing a hard nut, all repaired [53]. The Chinese consensus cites CBCT and printed guides as a means of reducing the risk of root perforation in obliterated teeth [22, 53].

7.7. Endodontics and periapical lesions

The multiple periapical abscesses of DGI result from pulp strangulation by obliteration or from pulp exposure by wear, and their risk justifies periodic radiographs [24]. Obliteration makes endodontics difficult or impossible, and attempting to negotiate obliterated canals risks lateral perforation in poorly mineralised dentine; apical surgery may be needed to retain an abscessed tooth [2, 24]. The Chinese consensus accepts that, when periapical lesions appear after obliteration, it is not always necessary to negotiate the canal, and that prolonged monitoring or periapical surgery are alternatives; pulp therapy is recommended only when the canals are identifiable on radiographs [22]. In DD-I, the AAPD described endodontic success in teeth without extremely short roots and short-term success of periapical curettage with retrograde filling in short roots, whereas Barron and colleagues do not recommend curettage with retrograde filling in short roots and Putrino and colleagues consider endodontics inappropriate in DD-I and periapical surgery reserved for long-rooted teeth — a divergence that reflects the lack of comparative series [2, 24, 44]. In DD-II, even shallow occlusal restorations may cause pulp necrosis, because the vascular channels of the pulp reach close to the dentine–enamel junction [24].

7.8. Bonding and materials

The dentine in DGI and DD has fewer tubules, of smaller diameter, and more peritubular collagen, which hampers adhesive infiltration and the formation of resin tags; the Chinese consensus recommends total-etch adhesive systems with rinsing (etch-and-rinse), which gave higher bond strength than self-etch systems [22]. Zhang and colleagues used glass ionomer cement, which performed similarly to resin cements in cast restorations in DGI, and attributed the decementations to severe wear and lower bond strength to the malformed tissue [53].

7.9. Orthodontics

Orthodontics is possible but requires prior assessment in patients with short roots or mobility, as in DD-I, and the use of light forces to reduce the risk of root resorption and exfoliation [22]. In DGI, in which the teeth are less hard, the Chinese consensus prefers clear aligners to brackets, to reduce stresses on the teeth, and the GeneReviews chapter on OI considers them a promising option [22, 48]. In the family of Putrino and colleagues, the proband had undergone rapid maxillary expansion for one year without a diagnosis of DD-I, and the authors do not exclude that orthodontics accelerated the root resorption that led to loss of the lateral incisors; rapid expansion, whether tooth-borne or bone-borne, causes external root resorption mainly in the first molars and upper incisors [44]. In contrast, Papagiannis and colleagues treated an 11-year-old girl with DD-I and bilateral impacted upper canines by extraction of the lateral incisors, surgical exposure and traction of the canines with light forces, achieving stable occlusion after 3 years and 2 months and one year after retention; genetic analysis found no variant in SMOC2, VPS4B or SSUH2, and the authors accept that a milder variant may explain the good response, so the result cannot be generalised to all forms of DD-I [34, 47]. Baylor College of Medicine is conducting, in the USA and Canada, a study of clear aligners in 57 participants with OI, with the PAR (Peer Assessment Rating) index as outcome, as yet without results [69].

7.10. Type I dentine dysplasia

In DD-I, the goal is to retain the teeth for as long as possible, but the prognosis for long-term retention is poor because of the short roots and periapical lesions [24]. The AAPD considered the crown–root ratio so unfavourable that prosthetic rehabilitation (complete denture, overdenture, partial denture or implants) is the only practical approach [24]. Chen and colleagues list conventional endodontics, periapical curettage and a preventive regimen, extraction of teeth with pulp necrosis and abscess, an overdenture in the primary dentition to maintain the vertical dimension, and implants to be considered at around 18 years [19]. Barron and colleagues advise removable prostheses until growth is complete and implants from 18 years, often after ridge augmentation, because the absence of roots and early loss cause maxillomandibular atrophy [2]. In the family of Putrino and colleagues, the lost lateral incisors were replaced by a removable flexible nylon prosthesis with aesthetic clasps, with psychological support for implant planning [44].

7.11. Type II dentine dysplasia

Permanent teeth in DD-II with normal shape, size and support can receive full crowns when necessary, and discoloured anterior teeth can be improved with resin, veneers or aesthetic crowns; if periapical lesions appear, the plan depends on root length [24]. The need for prosthodontic treatment is lower than in DGI, in line with the near-normal permanent dentition [23].

7.12. Dentinogenesis imperfecta in osteogenesis imperfecta

In OI, although there is no specific guideline, GeneReviews recommends full-coverage restorations in the primary dentition (stainless steel or zirconia crowns) and in the permanent dentition, and advises against intracoronal restorations, which favour loss of tooth structure [48]. Simple extractions are possible, but not immediately before or after an intravenous bisphosphonate infusion; oral surgery should, when possible, be performed before starting bisphosphonates or after stopping them, without resuming them before healing; patients on bisphosphonates should be monitored by a dentist; orthognathic surgery is discouraged in severe forms; and anxiolysis with nitrous oxide or midazolam can be used in the outpatient setting [48]. The AAPD recommended caution with procedures that transmit force to the jaws and with some forms of protective stabilisation, because of the risk of fracture, and described class III malocclusion with posterior crossbites and open bites in type I DGI [24]. Bisphosphonates influence dental development: in a pilot study of 58 patients (38 treated with intravenous pamidronate), enamel hypoplasia occurred only in those treated, more often when treatment began before 2 years of age, root dilaceration was associated with the number of cycles and pulp obliteration with the duration of treatment [70]; in another study, of 20 children treated for at least one year and 367 controls, bisphosphonates were associated with delayed root resorption and exfoliation of the primary molars, with no difference between children with and without DGI [71].

7.13. Psychosocial dimension

Developmental dental anomalies may be associated with low self-esteem, social avoidance, anxiety and depression, especially in adolescence, and the Chinese consensus recommends psychological interventions (cognitive behavioural therapy, family interventions, social skills training) as an adjunct [22]. The AAPD stressed behaviour guidance and psychological health at each stage [24]. In 2024 Toulouse University Hospital completed a study of 87 young people with AI or DGI on the psychosocial impact measured by the Child Oral Health Impact Profile (COHIP), whose results are not reported in the registry [69]. The German guideline notes that repeated costs throughout life constitute a barrier to access and that not all interventions are reimbursed, in contradiction with the United Nations Convention on the Rights of Persons with Disabilities [21].

7.14. Drugs, regulatory status and trials

There is no authorised medicine, nor any in clinical development, for DGI or DD; the medication used is that in routine use in dentistry (topical fluoride, local and general anaesthesia, anxiolysis), with no specificity for these disorders [15, 69].

None of these studies is an efficacy trial of an intervention in isolated DGI; the available treatment data come from case reports and case series, without control groups or standardised outcomes, which the German guideline acknowledges by grading the evidence as very low [21]. In the last 12 months, the developments have been diagnostic and classificatory, not therapeutic: the demonstration, in a French cohort, that long-read sequencing resolves exon 5 of DSPP (published observational study) [20]; the systematic review of genotype–phenotype correlation (published) [23]; the Chinese consensus and the English summary of the German S3 guideline (published consensus document and guideline) [21, 22]; the second family with the SSUH2 variant (published report) [46]; and the studies on the effect of bisphosphonates on the dentition in OI (published observational studies) [70, 71]. Work with dental organoids and dental pulp stem cells remains preclinical and does not translate into a clinical indication.

7.15. Pregnancy, perioperative care and transition

The sources consulted do not describe obstetric risks specific to isolated DGI or DD; in OI, there are specific obstetric and anaesthetic considerations [48]. In children, extensive treatment under general anaesthesia is an accepted option when cooperation does not allow phased treatment [21, 22]. The transition from paediatric to adult care coincides with the phase in which the German guideline places definitive rehabilitation (end of the second and beginning of the third decade), and fixed prostheses and implants are deferred until growth is complete [2, 21, 22].

7.16. Divergences between sources

The German guideline recommends rehabilitation with indirect crowns and full oral rehabilitation from the young permanent dentition, whereas Barron and colleagues advocate minimal preparation until adulthood and only then full rehabilitation [2, 21]. The AAPD accepted conventional restorative techniques in mild to moderate DGI and reserved full crowns for severe cases, whereas the Chinese consensus proposes preventive crowns in the primary dentition even with little wear [22, 24]. In DD-I, the AAPD and Barron and colleagues disagree on periapical surgery in short roots, and Putrino and colleagues consider endodontics inappropriate [2, 24, 44]. On caries, the Oxford Desk Reference considers that teeth with DGI are not more susceptible, whereas GeneReviews describes DGI as an independent risk factor in OI [48, 54]. None of these divergences is based on comparative studies.

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

The DSPP disorders are autosomal dominant with practically complete penetrance: each child of an affected person has a 50% probability of inheriting the variant, and the diagnosis is often evident in one of the parents [2, 15, 20]. Variable expressivity means that a child may have a milder or more severe form than the parent, or a different pattern in the two dentitions, as in the Chinese family in which the child had type II DGI in the primary teeth, the mother normal-looking permanent teeth with thistle-tube chambers and the grandfather a picture similar to his granddaughter's [5, 20]. De novo variants are rare but do occur [28]. The Chinese consensus and the German guideline recommend genetic counselling once a genetic dental disorder has been confirmed [21, 22].

In DD-I, the risk depends on the gene: 50% for the children of an affected person in the dominant forms (VPS4B, SSUH2 and most classical families) and 25% for the siblings of an affected person in the recessive forms, such as the SMOC2 form described by Bloch-Zupan and by AlFawaz and the Moroccan family without an identified gene; in recessive families, consanguinity should be asked about [12, 47]. GeneGuidance stresses that assuming dominant inheritance because the diagnosis is "type I DD" gives the wrong risk to a family with a recessive SMOC2 form [47]. The Italian family of Putrino and colleagues, with the same SMOC2 variant in a transmission interpreted as dominant, shows that the genetic information of each family should take precedence over generalisation [44]. In the DGI of OI, counselling follows the rules for OI, including parental mosaicism, present in up to 16% of families according to GeneReviews [48].

8.2. Cascade testing and minors

Testing relatives with an uncertain phenotype clarifies who needs dental surveillance, and in OI GeneReviews recommends clarifying the genetic status of at-risk relatives to guide dental, auditory, ophthalmological and spinal assessments [21, 48]. Because the dental phenotype of DSPP is visible from eruption, molecular testing of a child serves mainly to confirm the aetiology, distinguish isolated DGI from OI and allow early planning; the German guideline requires genetic counselling and the consent of the individual tested [21].

8.3. Prenatal and preimplantation diagnosis

Once the variant has been identified, preimplantation genetic diagnosis may be discussed with the couple, according to the Chinese consensus, in accordance with local ethical regulations [15, 22]. Because this is a disorder confined to the dentition and treatable, the decision rests with the couple after counselling; the sources consulted do not describe cases in which these techniques have been used for isolated DGI or DD.

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency

The classical estimate of 1 in 6,000 to 8,000 births for isolated DGI comes from Witkop and the United States population in the 1950s; de La Dure-Molla and colleagues note that in 2015 it was the only epidemiological figure available and that in other countries there were only reports and small cohorts [10, 16, 40]. Orphanet gives 14.6 per 100,000 for DGI type 2 in Europe and 14.5 per 100,000 for DGI as a whole, validated figures but based on expert opinion, without a confidence interval or defined population; DGI type 3 has unknown prevalence and DD type II is below 1 per 1,000,000, with 19 families recorded [15]. Zhang and colleagues cite a prevalence of 0.0022% for DGI-II in a Swedish study [53]. For DD-I, the estimate of 1 per 100,000 is due to Kalk and colleagues (1998) and Orphanet records 1.0 per 100,000 in Europe; in 1994 O'Carroll and Duncan counted 190 published cases of DD, 128 of them DD-I [7, 18]. The SMOC2 form has a prevalence below 1 per 1,000,000, with three published families, and the VPS4B and SSUH2 forms one and two families, respectively [32, 46, 47]. Witkop considered dentine disorders more frequent than enamel disorders, with US data from the 1970s [10].

9.2. Founder populations and geographical distribution

The tri-racial Brandywine isolate, in southern Maryland and Washington, is the founder population for DGI-III, with an incidence of genetic dental disorders of 1 in 15 [2, 10]. Most published families with DSPP variants are Asian (China, Korea, Japan, Thailand, Taiwan), which Gilani and colleagues attribute both to regional variation and to publication bias; exon 2 variants have been described in American, Asian and European families (Sweden, Denmark, United Kingdom) and intron 2 variants mainly in European families [29]. Exon 5 variants are distributed across several continents [29]. The c.52G>T p.(Val18Phe) variant has been found in Chinese, Korean, Caucasian and French families, a hotspot pattern rather than a founder effect [20, 28]. All families with DD-I due to VPS4B and SSUH2 are Chinese, and those with SMOC2 Turkish, Pakistani and Italian [7, 32, 33, 44, 46].

9.3. Europe, Portugal and Brazil

In Europe, the largest genetic series is the French Strasbourg cohort (42 families, 2016–2024), integrated into the national O-Rares network and the TETECOU Filière [20, 72]. The OI cohorts from Sweden and Romania document the frequency of type I DGI [49, 63]. The sources consulted contain no data from Portugal on DGI or DD. In Brazil, Acevedo and colleagues studied in Brasília three families from Bahia with dominantly inherited DGI-II (28 individuals examined, 19 affected), without bone pathology, with periapical lesions in six affected individuals and extensive tooth loss in the adults; analysis of exons 1 to 4 of DSPP found only polymorphisms, and exon 5 was not studied, so the molecular cause remained unresolved [40]. An earlier Brazilian family with DGI had a COL1A1 variant and was reclassified as mild OI [40].

9.4. Networks and associations

The relevant European reference network is ERN CRANIO, which includes a group dedicated to orodental anomalies; OI belongs to ERN BOND [25, 73]. In France, the O-Rares reference centre in Strasbourg mainly treats oligodontia, AI and DGI and participates in the orodental group of ERN CRANIO [72]. The sources consulted do not identify patient associations dedicated to isolated DGI; for the DGI of OI, there are the Portuguese Osteogenesis Imperfecta Association (APOI) and the Osteogenesis Imperfecta Foundation in the USA [74, 75].

10. Illustrative Clinical Case

10.1. Severe dentinogenesis imperfecta in a 3-year-old child

In 2025 Wang and colleagues described a Chinese family in which the proband, a 3-year-old girl, had primary teeth with opalescent discoloration, severe attrition and enamel pitting; the panoramic radiograph showed thin dentine and greatly enlarged pulp cavities, the shell-tooth appearance of type III DGI [36]. Her mother had similar discoloration and attrition, but her panoramic radiograph showed short, tapering roots with obliterated canals, the adult picture of DGI-II; her father had normal teeth, and no family member had signs of OI or hearing loss [36]. Wear had caused multiple pulp exposures with pulpitis and apical periodontitis, and the child was treated under general anaesthesia with pulpotomies, pulpectomies and stainless steel crowns or composite strip crowns, with a follow-up radiograph three months later [36]. Exome sequencing and Sanger confirmation identified in the proband and her mother, but not her father, the heterozygous variant NM_014208.3:c.38C>A p.(Ala13Glu), in the signal peptide, absent from gnomAD; a second missense variant in the DPP domain, c.3240T>G p.(Asp1080Glu), was not considered causal, and there were no co-segregating COL1A1 or COL1A2 variants [35, 36]. In HEK293E cells, DSP carrying p.(Ala13Glu) was not secreted into the medium, remaining in the cell lysate [36]. The figures of the paper show the pedigree, the intraoral photographs before and after treatment and the panoramic radiographs (https://pmc.ncbi.nlm.nih.gov/articles/PMC11880773/figure/mgg370087-fig-0001/), the mother's picture (https://pmc.ncbi.nlm.nih.gov/articles/PMC11880773/figure/mgg370087-fig-0002/), the chromatograms (https://pmc.ncbi.nlm.nih.gov/articles/PMC11880773/figure/mgg370087-fig-0003/) and the secretion assay (https://pmc.ncbi.nlm.nih.gov/articles/PMC11880773/figure/mgg370087-fig-0004/) [36].

The case brings together several points of this article: a signal peptide variant associated with a severe phenotype, as predicted by the genotype–phenotype correlation; the coexistence, in the same family, of shell teeth in the child and obliteration in the adult, which supports reading DGI-III as an early phase or severe extreme of the same spectrum; and the need for early treatment under general anaesthesia when pulp exposures are already present [21–23]. The management described is that of the report and coincides with the recommendations of the German guideline for the primary dentition (preformed stainless steel crowns and treatment under general anaesthesia when necessary) [21].

10.2. Radicular dentine dysplasia with late diagnosis

Putrino and colleagues described a 12-year-old boy referred by his paediatrician for orthodontic assessment, with poor oral hygiene, an upper lateral incisor in crossbite, mobility of the upper lateral incisors and lower first premolars, and previous abscesses that his parents had attributed to the shedding of primary teeth; he had undergone rapid maxillary expansion for one year, two years earlier [44]. The panoramic radiograph showed short or absent permanent roots, conical apical constrictions and crescent-shaped pulp remnants parallel to the cementoenamel junction, and radiographs at 7 and 10 years already showed the disorder in both dentitions (https://pmc.ncbi.nlm.nih.gov/articles/PMC10456812/figure/medicina-59-01477-f003/) [44]. His 10-year-old sister, without mobility, had the same signs on a panoramic radiograph taken a year earlier by another dentist, and his father, with lost teeth replaced by prostheses, had obliteration, short roots and abscesses (https://pmc.ncbi.nlm.nih.gov/articles/PMC10456812/figure/medicina-59-01477-f008/); no one had ever raised the suspicion [44]. Sanger sequencing identified NM_001166412.2(SMOC2):c.84+1G>T in the father and the two children and not in the 4-year-old sister (https://pmc.ncbi.nlm.nih.gov/articles/PMC10456812/figure/medicina-59-01477-f001/) [44]. The patients were then seen every 3 or 4 months; the proband spontaneously lost his lateral incisors, which were replaced by a removable flexible nylon prosthesis, and the lower left first premolar began to become mobile [44]. The report illustrates the cost of diagnostic delay and the need for caution with orthodontics in DD-I; the dominant interpretation of the transmission contrasts with the recessive inheritance of the founder SMOC2 form [7, 44].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The functional and aesthetic prognosis of DGI depends mainly on age at diagnosis and on the speed and quality of treatment: with early diagnosis and continuous care, good aesthetics and function can be achieved and nutritional deficits and psychosocial distress reduced [2]. Without treatment, the dentition may wear down to the gingiva [24]. Restorations require lifelong maintenance: in the twins of Zhang and colleagues there were decementations at 20, 23 and 37 months and a fracture at 40 months [53]. In DD-I the prognosis for tooth retention is poor: in the VPS4B family exfoliation began between 16 and 20 years and two patients were edentulous after 45; in the SSUH2 family loss began at about 20 years and six patients were edentulous at 35 [8, 9, 24]. The maxillomandibular atrophy that follows early loss constrains implant rehabilitation [2, 19]. The isolated forms have no known systemic involvement [15]; in the DGI of OI, the prognosis is that of OI.

11.2. Surveillance

The 3- to 6-month interval of the Chinese consensus and the six-monthly interval of GeneReviews for OI with DGI are compatible; the former applies to developmental dental disorders in general and the latter to OI [22, 48]. The frequency of radiographs is not quantified in the sources consulted; the AAPD justifies them by the risk of periapical abscesses and Barron and colleagues by the progression of obliteration [2, 24].

12. Summary and Final Considerations

Isolated hereditary dentine defects form two families. The first, caused by heterozygous DSPP variants, comprises Shields type II DD, type II DGI and type III DGI, which genetics has shown to be grades of severity of a single autosomal dominant disorder with practically complete penetrance and variable expressivity; the mechanism is dominant-negative, through retention of the mutant protein in the endoplasmic reticulum, not haploinsufficiency, and signal peptide variants tend to give more severe forms than exon 5 frameshifts. The second, radicular DD, has normal crowns and short or absent roots, is genetically heterogeneous (recessive SMOC2, dominant VPS4B and SSUH2, other families without a gene) and leads to early tooth loss. Type I DGI belongs to OI and requires a search for bone fragility in any child with opalescent teeth. Diagnosis is clinical and radiographic, but molecular confirmation of DSPP requires technology capable of reading exon 5, preferably long-read sequencing, and a negative result from a standard panel does not exclude the disorder. There is no targeted therapy: treatment is dental, preventive and restorative, begins at eruption of the primary teeth with preformed crowns and intensive prevention, continues in adolescence with indirect restorations on the molars and culminates in full oral rehabilitation in early adulthood, with recommendations from the German guideline that rest on very low evidence but are consensual. Genetic counselling distinguishes the 50% risk of the dominant forms from the 25% risk for siblings in the recessive forms of DD-I. The main gaps are epidemiological — there are no recent population-based estimates and no Portuguese data — and therapeutic, with an absence of comparative treatment studies.

13. References

1. Shields ED, Bixler D, el-Kafrawy AM. A proposed classification for heritable human dentine defects with a description of a new entity. Arch Oral Biol. 1973;18(4):543-53. doi:10.1016/0003-9969(73)90075-7
2. Barron MJ, McDonnell ST, Mackie I, Dixon MJ. Hereditary dentine disorders: dentinogenesis imperfecta and dentine dysplasia. Orphanet J Rare Dis. 2008;3:31. doi:10.1186/1750-1172-3-31
3. Online Mendelian Inheritance in Man (OMIM). #125490 Dentinogenesis imperfecta 1; DGI1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/125490
4. Online Mendelian Inheritance in Man (OMIM). #125500 Dentinogenesis imperfecta, Shields type III. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/125500
5. Online Mendelian Inheritance in Man (OMIM). #125420 Dentin dysplasia, type II; DTDP2. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/125420
6. Online Mendelian Inheritance in Man (OMIM). #605594 Deafness, autosomal dominant 39, with dentinogenesis imperfecta 1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/605594
7. Online Mendelian Inheritance in Man (OMIM). #125400 Dentin dysplasia, type IA; DTDP1A. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/125400
8. Online Mendelian Inheritance in Man (OMIM). #621440 Dentin dysplasia, type IB; DTDP1B. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/621440
9. Online Mendelian Inheritance in Man (OMIM). #621559 Dentin dysplasia, type IC; DTDP1C. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/621559
10. de La Dure-Molla M, Philippe Fournier B, Berdal A. Isolated dentinogenesis imperfecta and dentin dysplasia: revision of the classification. Eur J Hum Genet. 2015;23(4):445-51. doi:10.1038/ejhg.2014.159
11. Online Mendelian Inheritance in Man (OMIM). *125485 Dentin sialophosphoprotein; DSPP. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/125485
12. Cherkaoui Jaouad I, El Alloussi M, Laarabi FZ, Bouhouche A, Ameziane R, Sefiani A. Inhabitual autosomal recessive form of dentin dysplasia type I in a large consanguineous Moroccan family. Eur J Med Genet. 2013;56(8):442-4. doi:10.1016/j.ejmg.2013.05.003
13. Orphanet. Dentinogenesis imperfecta (ORPHA:49042). Available from: https://www.orpha.net/en/disease/detail/49042
14. Orphanet. Dentin dysplasia (ORPHA:1653). Available from: https://www.orpha.net/en/disease/detail/1653
15. GeneGuidance. DSPP — gene record and clinical curation (accessed 2026-09-26).
16. Witkop CJ. Hereditary defects in enamel and dentin. Acta Genet Stat Med. 1957;7(1):236-9. doi:10.1159/000150974
17. Xiao S, Yu C, Chou X, Yuan W, Wang Y, Bu L, et al. Dentinogenesis imperfecta 1 with or without progressive hearing loss is associated with distinct mutations in DSPP. Nat Genet. 2001;27(2):201-4. doi:10.1038/84848
18. GeneGuidance. VPS4B — gene record and clinical curation (accessed 2026-09-26).
19. Chen D, Li X, Lu F, Wang Y, Xiong F, Li Q. Dentin dysplasia type I—a dental disease with genetic heterogeneity. Oral Dis. 2019;25(2):439-46. doi:10.1111/odi.12861
20. Caravello G, Jiménez-Armijo A, Kawczynski M, Rey T, Antin M, et al. Diagnosis complexity of dentinogenesis imperfecta involving DSPP genetic variants. J Med Genet. 2026;63(6):344-51. doi:10.1136/jmg-2025-111345
21. 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(10):459. doi:10.1007/s00784-026-07142-w
22. Zhu J, Wan M, Duan X, Fan Z, Sun Y, Wang X, et al. Expert consensus on the diagnosis and management of tooth developmental anomalies. Int J Oral Sci. 2026;18(1):12. doi:10.1038/s41368-025-00401-8
23. Boonyakanog A, Sriwangyang K, Eamtanaporn T, Krasaesin A, Fakhruddin KS, Chantarangsu S, et al. Beyond the diagnosis: unraveling DSPP genotype-phenotype correlations in dentin dysplasia and dentinogenesis imperfecta. Jpn Dent Sci Rev. 2026;62:46-56. doi:10.1016/j.jdsr.2026.01.001
24. 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; revised 2013). Available from: https://www.aapd.org/assets/1/7/G_OHCHeritable2.PDF
25. European Reference Network CRANIO. Orodental anomalies. Available from: https://www.ern-cranio.eu/ (accessed 2026-10-29)
26. Zhang X, Zhao J, Li C, Gao S, Qiu C, Liu P, et al. DSPP mutation in dentinogenesis imperfecta Shields type II. Nat Genet. 2001;27(2):151-2. doi:10.1038/84765
27. Rajpar MH, Koch MJ, Davies RM, Mellody KT, Kielty CM, Dixon MJ. Mutation of the signal peptide region of the bicistronic gene DSPP affects translocation to the endoplasmic reticulum and results in defective dentine biomineralization. Hum Mol Genet. 2002;11(21):2559-65. doi:10.1093/hmg/11.21.2559
28. Kim JW, Hu JC, Lee JI, Moon SK, Kim YJ, Jang KT, et al. Mutational hot spot in the DSPP gene causing dentinogenesis imperfecta type II. Hum Genet. 2005;116(3):186-91. doi:10.1007/s00439-004-1223-6
29. Gilani M, Saikia A, Anthonappa R. The genetics of non-syndromic dentinogenesis imperfecta: a systematic review. Eur Arch Paediatr Dent. 2025;26(1):3-16. doi:10.1007/s40368-024-00992-6
30. Kim JW, Simmer JP. Hereditary dentin defects. J Dent Res. 2007;86(5):392-9. doi:10.1177/154405910708600502
31. Bloch-Zupan A, Jamet X, Etard C, Laugel V, Muller J, Geoffroy V, et al. Homozygosity mapping and candidate prioritization identify mutations, missed by whole-exome sequencing, in SMOC2, causing major dental developmental defects. Am J Hum Genet. 2011;89(6):773-81. doi:10.1016/j.ajhg.2011.11.002
32. Yang Q, Chen D, Xiong F, Chen D, Liu C, Liu Y, et al. A splicing mutation in VPS4B causes dentin dysplasia I. J Med Genet. 2016;53(9):624-33. doi:10.1136/jmedgenet-2015-103619
33. Xiong F, Ji Z, Liu Y, Zhang Y, Hu L, Yang Q, et al. Mutation in SSUH2 causes autosomal-dominant dentin dysplasia type I. Hum Mutat. 2017;38(1):95-104. doi:10.1002/humu.23130
34. Papagiannis A, Fanourakis G, Mitsea A, Karayianni K, Vastardis H, Sifakakis I. Orthodontic treatment of a patient with dentin dysplasia type I and bilateral maxillary canine impaction: case presentation and a family-based genetic analysis. Children (Basel). 2021;8(6):519. doi:10.3390/children8060519
35. 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
36. Wang Y, Xu X, Ding Y, Yuan G. A novel variant in dentin sialophosphoprotein (DSPP) gene causes dentinogenesis imperfecta type III: case report. Mol Genet Genomic Med. 2025;13(3):e70087. doi:10.1002/mgg3.70087
37. McKnight DA, Suzanne Hart P, Hart TC, Hartsfield JK, Wilson A, Wright JT, et al. A comprehensive analysis of normal variation and disease-causing mutations in the human DSPP gene. Hum Mutat. 2008;29(12):1392-404. doi:10.1002/humu.20783
38. National Center for Biotechnology Information. ClinVar: VCV000016856 (DSPP c.52G>T), VCV000016858 (c.202A>T), VCV000016853 (c.133C>T); VPS4B 4538522. Available from: https://www.ncbi.nlm.nih.gov/clinvar/ (accessed 2026-10-29)
39. Zhang Z, Chen Z, Huang J, Chen F, Dong Y, Gao X, et al. Splicing mutation in DSPP causes dentinogenesis imperfecta and amelogenesis imperfecta. BMC Oral Health. 2026;26(1):570. doi:10.1186/s12903-026-07872-z
40. Acevedo AC, Santos LJ, Paula LM, Dong J, MacDougall M. Phenotype characterization and DSPP mutational analysis of three Brazilian dentinogenesis imperfecta type II families. Cells Tissues Organs. 2009;189(1-4):230-6. doi:10.1159/000152917
41. Song YL, Wang CN, Fan MW, Su B, Bian Z. Dentin phosphoprotein frameshift mutations in hereditary dentin disorders and their variation patterns in normal human population. J Med Genet. 2008;45(7):457-64. doi:10.1136/jmg.2007.056911
42. Online Mendelian Inheritance in Man (OMIM). *607223 SPARC-related modular calcium-binding protein 2; SMOC2. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/607223
43. Alfawaz S, Fong F, Plagnol V, Wong FS, Fearne J, Kelsell DP. Recessive oligodontia linked to a homozygous loss-of-function mutation in the SMOC2 gene. Arch Oral Biol. 2013;58(5):462-6. doi:10.1016/j.archoralbio.2012.12.008
44. Putrino A, Caputo M, Galeotti A, Marinelli E, Zaami S. Type I dentin dysplasia: the literature review and case report of a family affected by misrecognition and late diagnosis. Medicina (Kaunas). 2023;59(8):1477. doi:10.3390/medicina59081477
45. Hu A, Lu T, Chen D, Huang J, Feng W, Li Y, et al. Vps4b heterozygous mice do not develop tooth defects that replicate human dentin dysplasia I. BMC Genet. 2019;20(1):7. doi:10.1186/s12863-018-0699-3
46. Deng Z, Cui Y, Peng L, Liu Z, Zeng B, Liu O, et al. [Clinical features and genetic mutation analysis of the hereditary dentin dysplasia type I family]. Hua Xi Kou Qiang Yi Xue Za Zhi. 2026;44(4):625-32. doi:10.7518/hxkq.2025184. PMID: 42576768
47. GeneGuidance. SMOC2 — gene record and clinical curation (accessed 2026-09-26).
48. Rodriguez Celin M, Steiner RD, Basel D. COL1A1- and COL1A2-Related Osteogenesis Imperfecta. 2005 Jan 28 [updated 2025 May 29]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1295/
49. Andersson K, Dahllöf G, Lindahl K, Kindmark A, Grigelioniene G, Åström E, et al. Mutations in COL1A1 and COL1A2 and dental aberrations in children and adolescents with osteogenesis imperfecta — a retrospective cohort study. PLoS One. 2017;12(5):e0176466. doi:10.1371/journal.pone.0176466
50. Lee J, Lee HR, Ko JM, Shin CH, Lee W, Song MH, et al. Genotype and phenotype of 198 Korean patients with nonlethal osteogenesis imperfecta: a retrospective cohort study. Hum Mutat. 2026;2026:5521427. doi:10.1155/humu/5521427
51. Online Mendelian Inheritance in Man (OMIM). #184260 Odontochondrodysplasia 1; ODCD1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/184260
52. 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: Oxford University Press; 2018. p. 160–162.
53. Zhang Y, Jin X, Zhang Z, Hu S, Jiang W, Pan H, et al. A novel approach to full-mouth rehabilitation of dentinogenesis imperfecta type II: case series with review of literature. Medicine (Baltimore). 2024;103(4):e36882. doi:10.1097/MD.0000000000036882
54. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 331–332.
55. 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: Oxford University Press; 2018. p. 501–505.
56. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 682–687.
57. Online Mendelian Inheritance in Man (OMIM). #166200 Osteogenesis imperfecta, type I; OI1. Baltimore (MD): Johns Hopkins University. Available from: https://www.omim.org/entry/166200
58. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 158–160.
59. 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: Oxford University Press; 2018. p. 497–500.
60. 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: Oxford University Press; 2018. p. 511–516.
61. 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: Oxford University Press; 2018. p. 517–520.
62. Malmgren B, Norgren S. Dental aberrations in children and adolescents with osteogenesis imperfecta. Acta Odontol Scand. 2002;60(2):65-71. doi:10.1080/000163502753509446
63. Dinulescu A, Stăiculescu ME, Dijmărescu I, Pavelescu ML, Păcurar D, Ulici A. Phenotypic variability and diagnostic characteristics of pediatric osteogenesis imperfecta: a 10-year multicenter cohort study from three tertiary pediatric hospitals in Bucharest, Romania. Diagnostics (Basel). 2026;16(13):1997. doi:10.3390/diagnostics16131997
64. 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: Oxford University Press; 2018. p. 548–551.
65. Chu KY, Prakash R, Wang SK, Simmer JP, Wright JT, Hu JC. Inherited dentin defects with comorbidities. J Am Dent Assoc. 2026;157:773-786.e1. doi:10.1016/j.adaj.2026.03.031
66. Lee K, Kramer K, Foster BL, Meyer BD. Dental manifestations of rare skeletal disorders: diagnosis and treatment for the oral health care provider. J Am Dent Assoc. 2026;157:904-16.
67. 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: Oxford University Press; 2018. p. 529–531.
68. 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: Oxford University Press; 2018. p. 543–544.
69. ClinicalTrials.gov. Records NCT02397824, NCT03810859, NCT04927962, NCT02934451, NCT04815564. Available from: https://clinicaltrials.gov/ (accessed 2026-10-29)
70. Demir K, Kardelen AD, Aslanger A, Koç C, Öztürk AP, Güleç Ç, et al. Evaluating dental developmental abnormalities in osteogenesis imperfecta: associations with timing and cumulative exposure to intravenous pamidronate — a pilot study. Clin Oral Investig. 2026. doi:10.1007/s00784-026-07194-y
71. Garcete Delvalle CS, De Nova García MJ, Mourelle Martínez MR. Exfoliation of primary dentition in children with osteogenesis imperfecta medicated with bisphosphonates. Eur J Paediatr Dent. 2026;27(1):24-30. doi:10.23804/ejpd.2025.2078. PMID: 41427456
72. Centre de référence des maladies rares orales et dentaires (O-Rares), Hôpitaux Universitaires de Strasbourg. Available from: https://www.o-rares.com/ (accessed 2026-10-29)
73. European Reference Network on Rare Bone Diseases (ERN BOND). Available from: https://ernbond.eu/ (accessed 2026-10-29)
74. Associação Portuguesa de Osteogénese Imperfeita (APOI). Available from: https://www.apoi.pt/ (accessed 2026-10-29)
75. Osteogenesis Imperfecta Foundation. Available from: https://oif.org/ (accessed 2026-10-29)