Mucolipidoses Types II and III (GNPTAB, GNPTG)

Autosomal recessive lysosomal diseases caused by deficiency of GlcNAc-1-phosphotransferase, which prevents formation of the mannose-6-phosphate marker: mucolipidosis (ML) II alpha/beta (I-cell disease), intermediate phenotypes and ML III alpha/beta (GNPTAB), and ML III gamma (GNPTG); persistent stuttering associated with the same genes is treated as an allelic association

Literature and regulatory information updated: 28 October 2026.

0. Initial Reference Table

1. Introduction

1.1 Definition and scope

Mucolipidoses II and III form a class of lysosomal storage diseases in which the affected enzyme is not a degradative hydrolase, but the enzyme that synthesises the signal targeting hydrolases to the lysosome. UDP-N-acetylglucosamine:lysosomal enzyme N-acetylglucosamine-1-phosphotransferase (GlcNAc-1-phosphotransferase, EC 2.7.8.17) is an α2β2γ2 hexamer whose alpha and beta subunits derive from a common precursor encoded by GNPTAB and whose gamma subunit is encoded by GNPTG; its deficiency prevents formation of mannose-6-phosphate (M6P) on newly synthesised hydrolases, which are then secreted instead of being delivered to the lysosome [1, 7, 31]. The result is a simultaneous intralysosomal deficiency of many enzymes in mesenchymal cells, with increased activity of those same enzymes in plasma and other fluids [7, 40]. Because the three classical phenotypes result from the same enzyme, the GNPTG GeneReviews chapter proposes treating them as “UDP-GlcNAc-1-phosphotransferase deficiency disorders” [7].

The scope of this article is the whole class. Mucolipidosis (ML) II alpha/beta is the severe form, with complete loss of activity, present at birth and fatal in early childhood in most cases [1, 25]. ML III alpha/beta is the attenuated form with residual activity between 1% and 10%, evident at around 3 years and compatible with survival into adulthood [1]. Between the two there is a continuum of intermediate phenotypes, one of which, associated with c.10A>C p.(Lys4Gln), constitutes a homogeneous clinical entity [1, 4]. ML III gamma, caused by biallelic variants in GNPTG, is generally milder than ML III alpha/beta and almost exclusively osteoarticular [7]. ML III gamma is grouped with the GNPTAB-related entities because it belongs to the same enzymatic class. Persistent non-syndromic stuttering, associated with different missense variants in the same genes and in NAGPA, is described as an allelic association, without a storage phenotype [38, 39]. The mucopolysaccharidoses (MPS), the skeletal lysosomal diseases, alpha-mannosidosis and mucolipidosis IV (gene MCOLN1) enter only into the differential diagnosis, the latter because it is a disease of a lysosomal cation channel unrelated to the M6P pathway [41].

1.2 Original description and historical overview

In 1966, Maroteaux and Lamy delineated clinically and radiologically a multisystem disease with progressive joint stiffness which they called “pseudo-Hurler polydystrophy”, because it shared with MPS I only the stiffness and dysostosis, without mucopolysacchariduria [1, 42]. In 1967, Leroy and DeMars, culturing fibroblasts from two children with Hurler features but without corneal clouding or excess urinary acid mucopolysaccharides, observed by phase-contrast microscopy dark, dense granules filling almost the entire cytoplasm, with markedly reduced beta-glucuronidase and increased acid phosphatase; they called them inclusion cells, abbreviated as I-cells [1, 43, 44]. In 1970, Spranger included I-cell disease, under the designation ML II, in the first clinical classification of the mucolipidosis group, positioned between the lipidoses and the MPS; biochemists considered the term inappropriate because “mucolipids” do not exist in nature, and the urinary excretion of oligosaccharides documented in the 1970s led to the alternative designation of oligosaccharidosis or glycoproteinosis [1]. The reference clinical description of I-cell disease was published in 1971 by Leroy, Spranger, Feingold, Opitz and Crocker [45].

The attenuated form had already been described before it had a name of its own. In Freiburg, Schinz and Furtwängler (1928) described, in a sibship of 11 children of first cousins, four siblings with generalised joint stiffness, flexion contractures of the fingers and toes, normal intelligence and destruction or abnormal development of the carpal and tarsal bones; Langer and colleagues (1966) reported a 61-year-old man with joint, carpal, tarsal and corneal changes, without urinary excess of acid mucopolysaccharides, but with excess of a glycoprotein [16]. Of the four cases in the original series of Maroteaux and Lamy, a pair of siblings later proved to have a form of MPS VI (Maroteaux–Lamy syndrome), and, of 18 patients with the pseudo-Hurler polydystrophy phenotype studied by Kelly and colleagues (1975), only 12 met the biochemical and ultrastructural criteria for ML III; the clinical phenotype alone does not define the disease [16]. The name “I-cell disease” as the designation of the clinical entity was given by Leroy and colleagues in 1969 [15].

1.3 Biochemical and molecular milestones

In 1972, Hickman and Neufeld proposed that the hydrolases of patients with I-cell disease were defective because they lacked the signal that allows their uptake and entry into lysosomes, which explained their accumulation in the culture medium and serum [40, 46]. In 1981, Reitman, Varki and Kornfeld demonstrated GlcNAc-1-phosphotransferase deficiency in fibroblasts from patients with I-cell disease and pseudo-Hurler polydystrophy: five ML II lines and five pseudo-Hurler lines transferred less than 0.02 pmol/mg protein/h, against 0.67–1.46 in controls, and another five pseudo-Hurler lines had 0.02–0.27, with normal or increased phosphodiesterase [47]. In the same year the two-step pathway was described: en bloc transfer of GlcNAc-1-phosphate linked in alpha to the 6-hydroxyl of mannoses of high-mannose oligosaccharides, followed by removal of the GlcNAc by a phosphodiesterase, the “uncovering enzyme”, which exposes the M6P [40, 48]. Varki and colleagues also identified an ML III variant with a catalytically normal enzyme when tested with alpha-methylmannoside as acceptor, but unable to recognise endogenous hydrolases; this variant corresponds to what would come to be called ML III gamma [40, 49]. Somatic complementation studies defined groups A (the most common, which includes many patients with ML II), B (rare) and C (uncommon, with normal activity on alpha-methylmannoside) [40]. By cell fusion, Honey and colleagues (1981) and Shows and colleagues (1982) defined two complementation groups in ML II and three in ML III; Ben-Yoseph and colleagues (1992) showed that fibroblasts from the variant forms of ML III (group C) phosphorylate mono- and oligosaccharides normally, but not hydrolases, and that a variant form of ML II belonged to the same group as the classical forms, which led to the proposal of a recognition subunit distinct from the catalytic subunit [8, 15].

The molecular basis of variant C was the first to be discovered: in 2000, Raas-Rothschild and colleagues cloned the human cDNA of the gamma subunit, mapped the gene to 16p, linked ML IIIC to this region in a large multiplex Druze family and identified a frameshift at codon 167 in three families [50]. In 2005, Tiede and colleagues, using recombinant gamma subunit as an affinity matrix, identified the product of GNPTA (then MGC4170), a 1256-amino-acid protein with two transmembrane domains, and found seven premature termination variants in six patients with ML II; retroviral transduction corrected the hypersecretion of lysosomal enzymes [51]. In 2006, Kudo, Brem and Canfield demonstrated that ML II and ML IIIA are allelic disorders of GNPTAB, with 15 variants in 18 lineages and both alleles identified in all [52]. The localisation of the gene followed an irregular path: linkage studies, somatic cell hybrids and gene dosage assigned the locus to 4q21-q23 (Mueller and colleagues, 1987); Canfield and colleagues (1998) placed it on 12p and showed that the 6.2 kb alpha/beta transcript was absent in 4 of 4 patients with ML II and markedly reduced in 2 of 2 with ML IIIA, with the gamma transcript at a normal level; Tiede and colleagues (2005) mapped it by genomic sequence analysis to 12q23.3, and the cytoband recorded in OMIM is 12q23.2 [11, 15]. With the cloning of the two genes, GlcNAc-1-phosphotransferase became an exception to the “one gene, one enzyme” principle [11]. In 2010, Kang and colleagues associated missense variants in GNPTAB, GNPTG and NAGPA with persistent non-syndromic stuttering [38].

1.4 Nomenclature and current classifications

In 2008, Cathey and colleagues proposed a nomenclature based on the gene and the affected subunit, which is the one used by GeneReviews and by variant reviews: ML II became ML II alpha/beta, ML IIIA became ML III alpha/beta and ML IIIC became ML III gamma [6, 40, 53]. The designation “Pacman dysplasia”, formerly given to a perinatal lethal skeletal dysplasia with epiphyseal stippling, corresponds, in most published cases, to the prenatal manifestation of ML II alpha/beta [1, 3, 40]. Wilcox and colleagues (2005) argued that Pacman dysplasia (MIM 167220) is distinct from ML II, but recognised that radiological and morphological criteria do not distinguish them; definitive diagnosis requires demonstration of lysosomal storage in pathological material or enzyme assay [15]. The intermediate phenotype associated with c.10A>C p.(Lys4Gln) was given the name “Webb-type mucolipidosis” [1, 4]. The designation ML I, which belonged to sialidosis (alpha-neuraminidase deficiency, gene NEU1), and ML IV, caused by MCOLN1, do not belong to this class; the term “mucolipidosis” thus covers four different inborn errors, of which only those linked to GNPTAB and GNPTG derive from the M6P pathway [1, 31]. In the classification of Inborn Metabolic Diseases (Saudubray, 7th edition), ML II and ML III are described with the oligosaccharidoses for convenience and, from a cellular perspective, ML II is placed among the defects of intracellular trafficking [31, 54]. OMIM maintains three phenotype entries, 252500 (ML II alpha/beta, with the synonyms I-cell disease and ICD), 252600 (ML III alpha/beta, which includes the cardiac-predominant “atypical ML III alpha/beta” described by Steet) and 252605 (ML III gamma, also designated ML IIIC, ML III complementation group C or Iranian variant form), and two gene entries, 607840 (GNPTAB) and 607838 (GNPTG); it does not assign a number of its own to the intermediate phenotype [8, 11, 13, 15, 16].

The clinical classification into ML II, intermediate and ML III rests on age at onset, severity and trajectory, and not on the location of the variants; the Bargal series (24 patients) concluded that there is an ML III–ML II clinical continuum and that classification should be based on these clinical criteria [55]. The genetic classification into alpha/beta and gamma has practical consequences: bone biopsies and urinary resorption markers show altered bone remodelling in patients with GNPTAB variants, but not in patients with GNPTG variants, and the authors recommend classifying ML III also by gene to guide antiresorptive therapy [56].

1.5 Table of the class

2. Molecular Genetics and Pathophysiology

2.1 The mannose-6-phosphate pathway and GlcNAc-1-phosphotransferase

Soluble lysosomal hydrolases are glycoproteins synthesised in the endoplasmic reticulum with high-mannose oligosaccharides linked to asparagine. In the cis-Golgi, GlcNAc-1-phosphotransferase transfers GlcNAc-1-phosphate from the donor UDP-GlcNAc to the 6-hydroxyl of selected mannoses; in the trans-Golgi, N-acetylglucosamine-1-phosphodiester alpha-N-acetylglucosaminidase (NAGPA, the “uncovering enzyme”) removes the GlcNAc and exposes the M6P; the M6P receptors then bind the hydrolases and route them to endosomes and lysosomes [40, 48, 60]. The selectivity of the pathway depends on the phosphotransferase recognising a protein determinant common to the hydrolases, and not only the oligosaccharide, which explains why the group C variant retains catalysis on alpha-methylmannoside and loses phosphorylation of the hydrolases [40, 49].

GNPTAB (12q23.2) encodes the alpha and beta subunits in a single transcript of about 6.2 kb; the alpha/beta precursor protein, of 1256 amino acids with two transmembrane domains, is cleaved in the Golgi, so that the alpha and beta subunits correspond to the amino and carboxyl ends of the precursor [1, 51]. Cleavage, at a lysine–aspartate bond, generates an alpha subunit of 928 amino acids and a beta subunit of 328, and is carried out by site-1 protease (S1P, encoded by MBTPS1), the same protease that activates the sterol regulatory element-binding proteins; cells lacking S1P do not activate the precursor and reproduce an ML II cellular phenotype [11]. The gene has 21 exons spread over about 85 kb, and the protein has 19 potential N-glycosylation sites and a modular structure with at least six domains, including two domains similar to Notch repeats and a domain similar to the DMAP1-binding domain [11]. The alpha and beta subunits are anchored to the membrane and contain the catalytic centre; the gamma subunit, soluble and encoded by GNPTG (16p13.3), binds directly to the alpha subunit and increases the activity of the enzyme for phosphorylation of specific hydrolases [7]. The mature gamma protein has 281 amino acids (about 31 kDa) and two N-glycosylation sites, and its 1.3 kb transcript is expressed in all tissues analysed [13]. Most GNPTG missense variants lie in the M6P receptor homology domain (MRH domain), which binds phosphorylated and non-phosphorylated high-mannose N-glycans of acceptor hydrolases and presents them to the catalytic subunits [7, 12]. The 3.4 Å cryo-electron microscopy structure of a hyperactive truncated form of the enzyme (S1S3), published in 2024, increases phosphorylation of secreted enzymes and their uptake and is of interest for producing recombinant enzymes with more M6P [61]. The N-terminal cytoplasmic tail of the alpha/beta precursor, where p.Lys4Gln is located, is required for retention of the enzyme in the Golgi; the substitution maintains almost normal catalytic activity in vitro, but compromises Golgi retention of the enzyme [1, 62].

2.2 Cellular, tissue and physiological mechanism

Without M6P, newly synthesised hydrolases follow the default secretory pathway: the activity of almost all hydrolases rises 5 to 20 times (and, for some, up to 100 times) in plasma and culture medium, whereas in mesenchymal cells, such as fibroblasts, intracellular activity is low and undegraded macromolecules accumulate, the origin of the inclusions that gave the disease its name [7, 40, 44]. Some cell types use M6P-independent pathways: acid beta-glucosidase (glucocerebrosidase) maintains normal activity in fibroblasts and, in ML II, hydrolase activity in leucocytes is normal [7, 40]. The 2026 version of the GNPTAB GeneReviews describes intracellular activities as “markedly elevated in leukocytes and fibroblasts”, and the Turkish series by the same authors reports increased activities of beta-hexosaminidase, alpha-mannosidase and alpha-fucosidase in leucocytes and dried blood spots [1, 63]. For fibroblasts, this wording contradicts the mechanism and the other sources, which describe low intracellular activity (GNPTG GeneReviews, Atlas of Inherited Metabolic Diseases, Bone Dysplasias, Smith's Recognizable Patterns); for leucocytes, the GNPTG GeneReviews indicates normal activity in ML II and states that ML III gamma is not diagnosed by leucocyte hydrolases [2, 7, 40, 44]. This article adopts the description on which most sources agree (activity elevated in plasma and low in fibroblasts) and considers that the leucocyte result does not, by itself, serve to confirm or exclude the disease. The lack of hydrolases in the lysosome leads to accumulation of cholesterol, phospholipids and glycosaminoglycans (GAGs) [60]. There is also secondary deficiency of sialidase (neuraminidase) in fibroblasts and leucocytes, a 3- to 4-fold increase in sialic acid in cultured cells and urinary excretion of a sialyl-hexasaccharide, and the hydrolases secreted by I-cells retain sialic acid residues that alter their electrophoretic mobility [15]. Leroy and colleagues (1972) found no accumulation of lipids or mucopolysaccharides in the brain and viscera, which is why they questioned the very designation “mucolipidosis” [15].

Tissues of mesenchymal origin concentrate the phenotype. The first autopsy of ML III alpha/beta, in a 45-year-old woman, showed cytoplasmic lysosomal granules in bone, cartilage, brain (neurons and glia), myocardium, cardiac valves and hepatic and cutaneous fibroblasts, with tracheobronchomalacia [64]. Excessive release of lysosomal glycoproteins into the extracellular matrix is probably the main cause of the interstitial lung disease of ML II, and slow degeneration of extracellular matrix connective tissue explains the decline of the lung parenchyma in ML III [1]. Progressive thickening of the airway mucosa, stiffness of the connective tissue and rib cage and deposition in the pulmonary interstitium lead to respiratory failure and pulmonary hypertension (PH) [1, 44].

In bone, ML II combines decreased bone formation and increased osteoclastogenesis, according to the murine model of Kollmann and colleagues cited by GeneReviews [1, 65]. Patients with GNPTAB variants (ML II and ML III alpha/beta) have increased bone resorption, measured by urinary collagen crosslinks, and altered remodelling on iliac crest biopsies, whereas patients with GNPTG variants and Gnptg-knockout mice do not; this difference is the proposed explanation for the milder phenotype of ML III gamma [7, 56]. In the perinatal period, ML II causes fractures and rickets-like and hyperparathyroid changes, related to transient secondary hyperparathyroidism that is possibly attributed to deficient calcium transport to the fetus [31, 66]. How the metabolic defect produces the central nervous system manifestations remains unclear [31].

Spontaneous and induced animal models exist. A cat with spontaneous mucolipidosis reproduces the facial dysmorphism, dysostosis multiplex and inclusion cells, with low hydrolases in fibroblasts and elevated in plasma and obligate carriers with intermediate serum values, but, unlike the human disease, develops diffuse retinal degeneration with blindness at 4 months [15]. The Nymphe mouse, obtained by chemical mutagenesis, has the nonsense variant p.Tyr867Ter in Gnptab, upstream of the cleavage site: it loses the beta subunit, retains the truncated alpha subunit in the endoplasmic reticulum and shows, from birth, facial and skeletal changes, with progressive ataxia due to loss of Purkinje cells, reduced fertility and increased mortality [11, 15].

2.3 Inheritance, penetrance and expressivity

All forms of the class are transmitted in an autosomal recessive manner; heterozygotes are asymptomatic and are not at increased risk of disease [1, 7]. Each patient’s phenotype depends mainly on the residual activity conferred by the less severe allele: two variants without functional activity produce ML II, and a variant with some activity in trans with a null variant “protects against the ML II phenotype” [1, 67]. Expressivity is variable even with the same genotype: in the Manchester series, five patients with ML II showed phenotypic variability with the same genotype, and GeneReviews mentions clinically heterogeneous homozygous genotypes [1, 68]. The Atlas of Inherited Metabolic Diseases notes that, in ML III, females may be taller and less affected [40]. Intrafamilial variability can be marked in ML II: among the 9 patients studied by Beck and colleagues (1995), a boy survived to 8 years and died of bronchopneumonia, whereas his two sisters died with a similar course, one of heart failure at 2 months and the other at 29 days [15]. A patient with an atypical mosaic form of ML II has been described, with one cell population carrying the defect and another normal, without evidence of twin chimerism [15]. In heterozygotes, intracellular hydrolases are normal, but GlcNAc-1-phosphotransferase activity is intermediate [15].

2.4 Spectrum of variants in GNPTAB and genotype–phenotype correlation

The review by Velho and colleagues (2019) brought together 564 individuals with ML II or ML III published since the identification of the genes, with 258 different variants in GNPTAB and 50 in GNPTG [53]. The systematic review by Dogterom and colleagues (2021), with 843 patients published between 1968 and August 2019, identified pathogenic variants in 388 patients (571 GNPTAB alleles and 179 GNPTG alleles); the most frequent variant in ML II was c.3503_3504del and, in ML III alpha/beta, c.22A>G; all pathogenic GNPTG variants resulted in ML III gamma, with no patient having a GNPTG variant and a second variant in GNPTAB [25]. Parents were consanguineous in 170 of 436 cases with that information (39.0%; 20.5% of the total), which explains homozygosity in about half of genotyped patients (189; 53.9% in ML II and 49.7% in ML III), and biallelic variants were known in 362 of 388 (93.3%) [25]. In homozygosity, nine GNPTAB variants always produced ML II — the nonsense variants c.136C>T, c.1090C>T and c.3565C>T, the frameshift variants c.3487_3490del, c.1314_1315del and c.3503_3504del (the latter in 47 patients), the missense c.242G>T and the splice variants c.3135+1G>T and c.3434+1G>A —, and c.3443_3446del behaved as a severe variant in trans with c.3503_3504del; by contrast, c.3173C>G, c.1120T>C, c.1196C>T, c.1208T>C, c.2345C>T and c.2715+1G>A produced ML III alpha/beta when combined with one of those severe variants; in GNPTG, the most frequently reported variant was c.323G>A [25]. These notations reproduce those of the article, which does not state the transcript; those listed in the variant table are normalised on the MANE transcript. More than 300 individuals with pathogenic variants in GNPTAB had been identified at the time of the GeneReviews revision [1].

In ML II, the variants are homozygous or compound heterozygous without functional activity, typically nonsense or frameshift [1]. In ML III alpha/beta, residual activity lies between 1% and 10% and the variants include missense and several splice variants, often in compound heterozygosity with a null allele [1]. Kudo and colleagues observed that splice variants in the invariant splice-site sequence are associated with ML II and those outside it with ML III alpha/beta [52]. The intron 17 example illustrates the rule: c.3335+1G>A, in the invariant sequence, and c.3335+6T>G, outside it, produce the same abnormal mRNA, with skipping of exon 17 and premature termination, but fibroblasts with the former, associated with ML II, have less than 0.1% activity, and those with the latter, associated with ML III alpha/beta, 1% to 3%; the location outside the invariant sequence is the proposed explanation for the higher activity [10, 11]. In Japan, in 25 patients with ML II and 15 with ML III, homozygosity or compound heterozygosity for nonsense and frameshift variants produced the severe phenotype, whereas p.(Phe374Leu), p.(Asn1153Ser) and splice variants produced the attenuated phenotype even in trans with a nonsense variant; the recurrent variants were p.(Gln104Ter), p.(Trp894Ter), p.(Arg1189Ter) and c.2715+1G>A (exon 13 skipping) [67]. In that series, p.(Arg1189Ter) accounted for 41% of the 80 alleles and p.(Phe374Leu) for 10%, with signs of a founder effect for the latter; no patient with ML III was homozygous for p.(Arg1189Ter), and most were compound heterozygous for p.(Arg1189Ter) and p.(Phe374Leu) [11]. The North American series of Cathey and colleagues (61 probands) described 51 variants, 42 novel, and associated ML II with an almost total absence of activity in leucocytes and ML III with at least one missense or splice allele [24]. The series of Bargal and colleagues (21 families with ML II and 3 with ML III, mainly from Israel, Palestine and Turkey) concluded that the phenotype depends on the severity and not the location of the variants [55].

The homogeneous intermediate form was defined by Leroy and colleagues (2014) in 8 patients with c.10A>C p.(Lys4Gln) in trans with a frameshift variant: phosphotransferase activity of 7–12%, with most of the newly synthesised cathepsin D retained inside the cell [1, 4]. At the other end of the spectrum, Steet and colleagues described a 47-year-old woman with dilated cardiomyopathy and mild sensory neuropathy, without skeletal anomalies, homozygous for a variant in the last nucleotide of exon 7 that causes exon skipping, with 3% activity in fibroblasts [57]. That variant, NM_024312.5:c.771G>A p.(Leu257=), was identified again in 2026, in compound heterozygosity with c.2502del p.(Lys834AsnfsTer5), in a 38-year-old man with arrhythmogenic left ventricular cardiomyopathy [10, 58]. GeneReviews mentions a patient with ML III homozygous for “c.342delCA”, born to consanguineous parents, with more marked intellectual disability than usual; the notation is not a valid description according to the Human Genome Variation Society (HGVS) and the two-nucleotide deletion in that region is repositioned by VariantValidator to c.344_345del p.(Thr115AsnfsTer5), without the bases having been confirmable in the primary publication [1, 10].

The following table brings together the GNPTAB variants of clinical or population interest, normalised in VariantValidator on the MANE Select transcript and, when available, with the ClinVar classification (consulted on 27 October 2026).

Population studies add regional alleles. In Brazil, in 12 unrelated patients with ML II or ML III alpha/beta, 9 variants were identified, among them c.242G>T p.(Trp81Leu), c.1208T>C p.(Ile403Thr), c.1514G>A p.(Cys505Tyr), c.1759C>T p.(Arg587Ter) and c.2269_2273del, and c.3503_3504del accounted for 11 of the 24 alleles, which led the authors to suggest starting the analysis with exon 19 [71]. In a series of 13 patients (10 Portuguese, 1 Finnish, 1 Spanish of Arab origin and 1 Indian), Encarnação and colleagues found 11 variants, 8 novel, including in GNPTAB c.121del, c.440del, c.2249_2250insA, c.242G>T, c.1208T>C and c.1999G>T, and observed by quantitative polymerase chain reaction (PCR) signs of feedback regulation between the alpha/beta and gamma subunits [76]. In that series, 8 of the 9 patients with ML II had a nonsense or frameshift variant, c.3503_3504del accounted for 45% of mutant alleles and one patient with ML II was homozygous for a missense variant; the three patients with an ML III phenotype had missense variants [11]. In Korea, all 7 patients were compound heterozygous, without a population hotspot [73]. In 2026, a series of 8 patients with ML II from Egypt and Yemen described three novel variants, among them c.3602+2T>C, and proposed haploid HAP1 cells with GNPTAB knocked out as a model for assessing pathogenicity [77]; an Iranian series of 6 unrelated patients with ML (homozygous variants in GNPTAB, GNPTG and MCOLN1) included two cases with c.3503_3504del [78].

2.5 Spectrum of variants in GNPTG

In GNPTG, correlation data are limited and most variants are loss-of-function; GeneReviews highlights c.347_349del, c.445del, c.499dup and c.607dup as recurrent [7]. In the Turkish series of Tüysüz and colleagues (11 patients from 8 families), the insertions c.493_494insC and c.283_284insC were associated with a severe phenotype and c.196C>T, c.347_349del, c.652_655del and the combination c.445del/c.367A>G with a milder phenotype, in the authors’ notation [26]. The deletion c.347_349del p.(Asn116del) is in-frame, so its effect cannot be deduced from the variant class; whereas GeneReviews includes it among the loss-of-function variants associated with a severe phenotype, the Tüysüz series associates it with a milder phenotype [7, 10, 26]. The missense variants c.316G>A p.(Gly106Ser), c.376G>A p.(Gly126Ser) and c.425G>A p.(Cys142Tyr) cause misfolding and retention of the gamma subunit in the endoplasmic reticulum, whereas c.857C>T p.(Thr286Met) does not alter function and is probably benign [79]. Three Afghan siblings, born to consanguineous parents and examined between 14 and 17 years, had begun at around 7 years with progressive joint pain and stiffness (fingers, hips and spine), with myopia, normal hearing and intelligence, short stature in the two girls and normal height in the boy; they were homozygous for the intronic variant IVS7-10G>A (historical numbering, = NM_032520.5:c.527-10G>A, p.(Val176GlyfsTer18) according to the authors), which creates a cryptic acceptor site, markedly reduces the mRNA by nonsense-mediated decay and abolishes the gamma subunit [13, 25, 80]. In the boy’s fibroblasts, intracellular hydrolases were at 30–50% of control values and about 70% of newly synthesised cathepsin D was secreted instead of being delivered to the lysosome [8, 80]. Persichetti and colleagues identified six novel variants in 10 patients from 7 families, among them c.611del p.(Gly204AlafsTer6) in homozygosity in two Italian siblings, c.333G>A p.(Trp111Ter) in homozygosity in two Spanish siblings, c.640_667del p.(Glu214LysfsTer37) in a Spanish boy and the combination of c.610-2A>G (which produces skipping of exon 9 or retention of introns 8 and 9) with c.316G>A p.(Gly106Ser) in two Italian siblings [10, 13, 81]. Encarnação’s Portuguese patient, compound heterozygous for c.639del and c.610-1G>T, had onset at 10 years, without psychomotor delay, and reached adulthood; the GNPTG and GNPTAB transcripts were reduced 10- and 2.4-fold, and the authors attributed the mild phenotype to the fact that the gamma subunit does not take part in catalysis [13, 76]. Two intronic deletions of a size detectable by sequencing may escape analysis because of their location [7, 81].

2.6 Founder variants and origin

The c.3503_3504del variant has been described in Muslim Arab patients from Israel and Palestine, Turkish, Canadian, Italian, Portuguese, Irish Traveller, North American, Argentinian and Brazilian patients; Coutinho and colleagues analysed 44 homozygous or compound heterozygous patients and 16 carriers and found a common haplotype (59 of the 61 mutant chromosomes, 97%, shared 4 of the 5 polymorphic markers, and the remaining two, from Italian patients, differed at a single marker), with an estimated age of the mutation of about 2063 years and a probable peri-Mediterranean origin [15, 70]. In Saguenay–Lac-Saint-Jean, genealogical reconstruction of 27 parents of 16 deceased children pointed to six founders (three couples) and the variant was present in all obligate carriers and absent in 50 controls [23]; GeneReviews places the introduction of the variant in the 17th century, by immigrants from France and Scotland, whereas OMIM, based on Plante, states that the six founders all originated from France and married in Quebec in the second half of the 17th century [1, 15]. In GNPTG, the c.499dup variant is a founder in the Druze and Bedouin of northern Israel, with a carrier frequency of 1/59 [7]. In China, c.1090C>T p.(Arg364Ter) was identified in 39.3% of 56 patients with ML II [72].

2.7 Allelic association: persistent non-syndromic stuttering

The GNPTAB and GNPTG GeneReviews chapters state that there are no other phenotypes associated with germline variants in these genes [1, 7], and OMIM records stuttering (STUT2, MIM 609261) in the GNPTAB and GNPTG entries as an association pending confirmation [11, 13]. The primary literature, however, describes an association with persistent stuttering. Kang and colleagues (2010) studied a large consanguineous Pakistani family with linkage to the 12q23.3 region and identified in GNPTAB the variant c.3598G>A p.(Glu1200Lys), present in about 10% of Pakistani families with stuttering, in 1 of 192 chromosomes from Pakistani controls and in none of 552 chromosomes from North American controls; two homozygous individuals were unaffected, reflecting incomplete penetrance [38]. Co-segregation was incomplete in both directions: 13 affected individuals were heterozygous and 12 homozygous, but 3 affected individuals were not carriers, and 2 homozygotes and 9 heterozygotes were unaffected [11]. Haplotype analysis of 8 carriers placed the origin of p.(Glu1200Lys) on a founder allele about 572 generations old, or about 14,300 years [11]. The same authors found, in cases and not in controls, another three variants in GNPTAB (c.961A>G p.(Ser321Gly), c.1363G>T p.(Ala455Ser) and c.1875C>G p.(Phe624Leu)), three in GNPTG and three in NAGPA [38]. In a sample of 1013 individuals with persistent stuttering, Raza and colleagues identified 164 carriers of rare non-synonymous variants in GNPTAB, GNPTG or NAGPA (81 different variants, almost all missense and only one previously described in mucolipidosis) and estimated that this pathway may be involved in up to 16% of cases, with variants at distinct sites and a less severe effect than that of mucolipidosis variants [39]. In mice, the human variants p.Ser321Gly and p.Ala455Ser produced deficits in the flow of ultrasonic vocalisations and a reduction of astrocytes in the corpus callosum, and only astrocyte-specific inactivation reproduced the deficit [82]; a mouse with the homologous substitution p.Glu1179Lys emitted fewer vocalisations per unit of time, with longer pauses and more stereotyped patterns, without differences in an extensive battery of non-vocal behaviours [11]; a 2025 preprint, not peer-reviewed, described changes in fine motor kinematics in a model with a stuttering-associated variant [83]. For counselling, this association does not equate to a phenotype in carriers of mucolipidosis variants: in ML III alpha/beta, language develops at the expected age and GeneReviews explicitly records the absence of stuttering [1].

3. Clinical Spectrum and Manifestations by System

3.1 Spectrum and data sources

The clinical spectrum of the class ranges from ML II with prenatal manifestations and death in early childhood to forms of ML III diagnosed in adulthood because of isolated cardiomyopathy. Natural history data come from series from specialised centres and from reviews of published cases, which are not population cohorts and over-represent severe and unusual cases. The following table summarises the quantitative sources used throughout the article.

3.2 Mucolipidosis II alpha/beta

3.2.1 Prenatal and neonatal period

ML II is evident at birth or even before. Suggestive neonatal findings include small to low-normal anthropometry for gestational age, restricted passive joint mobility (especially at the shoulders), a flat facial profile with shallow orbits and a depressed nasal bridge, thick skin with a waxy texture (most evident around the earlobes), chest deformity with kyphosis, clubfeet, deformed long bones, hip dislocation, gingival hypertrophy, umbilical or inguinal hernias, respiratory and feeding difficulty and, rarely, hydrops fetalis [1]. Birth weight is often below 2.5 kg [44]. In the prenatal setting, the disease may present as a skeletal dysplasia, with fractures, and has been described under the name Pacman dysplasia [3, 31, 88]; Michels and colleagues (1982) included ML II among the diseases that may manifest intrauterine fractures [15]. Coarse face at birth, craniosynostosis and limb shortening are neonatal signs that Inborn Metabolic Diseases associates, among other diseases, with I-cell disease [89]. Neonatal cholestasis and proximal tubular dysfunction are also described [44].

Some newborns show radiological changes of rickets and hyperparathyroidism, with or without biochemical secondary hyperparathyroidism, which begin in the neonatal period and regress by the first year [44, 66]. In the Irish series, 5 of 21 children had hyperparathyroidism, all with radiological signs (prenatal fractures in 3 and one acute fracture), and 4 with biochemical secondary hyperparathyroidism; 3 received high-dose vitamin D and 2 with fractures were handled as little as possible, and resolution occurred in all, regardless of treatment [22].

3.2.2 Growth and skeleton

Postnatal growth is limited and often ceases in the second year of life; measured length appears to “decrease” as hip and knee contractures worsen, and head circumference is proportionate to the body [1]. The range of all joints, large and small, is markedly limited; shoulder mobility is markedly reduced despite consistent axial and appendicular hypotonia; the wrists lose range and the hands and fingers broaden in the years after infancy, becoming stiff in palmar flexion, claw-like and deviated from the axis [1]. Dorsolumbar kyphosis, chest deformity, clubfeet, hip dislocation, broadening of wrists and fingers and abnormal curvatures of the long bones are described [44]. In the series of 16 patients of Ammer and colleagues, supra-acetabular constriction of the ilium was present in all, femoral cloaking in 7, rickets-like or hyperparathyroid changes in 6 and femoral neck constriction in 7; the acetabular and migration indices increased significantly in females, without relevant progression in the group as a whole (mean acetabular index 23.0° at the first and 23.7° at the last assessment) [84]. Cervical canal stenosis was found in 63% of the Hamburg patients who underwent magnetic resonance imaging (MRI) [85].

3.2.3 Face, skin and oral cavity

The face is flat at birth, with a depressed nasal bridge, shallow orbits and a prominent mouth; facial coarsening begins in early childhood and progresses, with a high, narrow forehead, thin eyebrows, puffy eyelids, persistent epicanthic folds, anteverted nares and a long philtrum [1, 44]. Gingival hypertrophy appears shortly after birth and is so marked that tooth eruption appears incomplete; the Atlas of Inherited Metabolic Diseases states that gingival hypertrophy is always seen in ML II [1, 40]. The skin is thick and tight in infancy, less tight with age; there are tortuous periorbital veins, telangiectasias on the cheeks, fine hair sometimes white to golden and, in some patients, cavernous haemangiomas [1, 44]. Metopic prominence leads to suspicion of craniosynostosis, rarely formally confirmed, and this suspicion has in some cases prompted inappropriate cranial surgery [1]. The systematic review of 340 patients with MPS or ML II/III described thick lips, midface hypoplasia, a high-arched palate, hypoplastic condyles, coronoid hyperplasia, macroglossia, gingival hyperplasia, thick dental follicles, dentigerous cysts, malformed teeth, enamel defects and open bite in all subtypes [90].

3.2.4 Heart

Cardiac involvement is probably present in all children with ML II: thickening and regurgitation of the mitral and, less frequently, aortic valve are the most common findings; right ventricular or global hypertrophy and PH have been described, but are rarely documented objectively; valvular changes are slowly progressive and rapidly progressive cardiomyopathy is uncommon [1, 91]. Inborn Metabolic Diseases mentions severe cardiomyopathy, valvular disease and coronary disease, and the OrphanAnesthesia recommendation includes rare coronary occlusion among the findings to look for [31, 92]. In China, cardiac involvement was recorded in 16.1% of 56 patients, whereas GeneReviews considers that cardiac involvement is probably present in all children [1, 72].

3.2.5 Airways and lung

The voice is hoarse and breathing noisy throughout life; the airways are narrow because of slowly progressive mucosal thickening and connective tissue stiffness, the parenchyma is affected and chest stiffness worsens restrictive insufficiency [1]. Obstructive sleep apnoea (OSA), bronchitis, pneumonia and otitis are recurrent [44]. PH is probably the rule in patients who survive into childhood and has been documented as severe in a long-term survivor [1, 91]. In Manchester, all documented overnight oximetry studies showed sleep-disordered breathing; all patients needed oxygen, some continuous positive airway pressure (CPAP) or bilevel pressure (BiPAP), 4 of 5 needed nasogastric tube feeding, and there were 2 difficult intubations, one of which ended in death [68]. Respiratory support is rarely needed in the newborn; OSA requires nocturnal support in some children and a minority of the longest survivors depend on persistent assisted ventilation, always started during treatment of an acute infection [1]. Pneumothorax was described in 2026 as a potentially fatal late complication [93].

3.2.6 Feeding, abdomen and other organs

Initial weak sucking resolves quickly; caloric requirements are lower than those of an unaffected sibling, because of short stature and low activity, and only a minority need gastrostomy [1]. Mild to moderate hepatomegaly is common and splenomegaly less frequent; umbilical hernias are almost constant, may enlarge and do not require repair, and inguinal hernias are slightly more frequent than in the average infant [1].

3.2.7 Neurodevelopment, hearing and vision

Most children with ML II have a near-normal level of alertness and express affection, joy and displeasure like any child [1, 24]. They sit with support at around 1 year and without support sometimes only at 2 years; most never walk independently; expressive language is limited to single words and comprehension, which is better, is not age-appropriate; intellectual disability is present in all, with variable severity [1]. Smith's Recognizable Patterns describes a developmental plateau at around 18 months without apparent subsequent deterioration [44]. In Hamburg, 32 neurocognitive assessments in 11 patients (median age 34 months) showed a mean developmental quotient of 36.7% (standard deviation 20.4) at the last assessment, with a mean gain of 0.28 age-equivalent points per month (confidence interval 0.17–0.38), and MRI showed non-specific, non-progressive changes (mild atrophy, white matter lesions); the authors concluded that there is profound impairment without neurodegeneration [85]. Recurrent otitis media is common and causes conductive deafness; severe and sensorineural deafness are less frequent; corneal opacity is typically mild, visible only on slit-lamp examination, and significant visual impairment is uncommon [1].

3.2.8 Course and causes of death

ML II is severe and progressive [1]. The most common cause of death is refractory cardiorespiratory failure, often after a respiratory infection [1, 44]. Estimates of age at death vary with the source: Smith's Recognizable Patterns states “usually by 5 years”, Inborn Metabolic Diseases the first decade, the review of 843 cases a median survival of 5.0 years and a median age at death of 1.8 years, the Turkish series a median survival of 28 months and the Manchester series a mean survival of 36 months [25, 31, 44, 63, 68]. These differences reflect the period, the method of data collection and the level of supportive care. In the review of 843 cases (516 with ML II and 313 with ML III, 14.0% published before 1980 and 65.1% after 2001, mainly from the USA, Japan, Italy and China, followed by Israel, Brazil and Turkey), symptom onset had a median of 0.0 years in ML II (interquartile range 0.0–0.3) and 3.0 years in ML III (2.0–6.0; up to 47 years), and age at diagnosis, 0.7 and 9.0 years, did not change over time [25]. The presenting complaints, described in 406 patients, were in ML II facial dysmorphism (47.4%), developmental delay (24.2%), bone changes (20.0%), growth retardation (12.6%), abnormal skull shape (10.7%) and joint limitation (10.7%), and in ML III hand deformities (claw hand, stiffness, carpal tunnel syndrome, inability to make a fist; 47.6%), joint limitation (30.9%) and growth retardation (13.3%) [25]. At the time of description, 166 of 516 patients with ML II (32%) and 12 of 313 with ML III (4%) had died; the cause, known in 93 deaths from ML II, was pneumonia in 32, respiratory failure in 24, heart failure in 15, both in 7, multiorgan failure in 6 and complications of stem cell transplantation in 3, and in the five deaths from ML III with a known cause, heart failure in 2, pneumonia, complication of heart transplantation and renal failure [25]. Adjusted for phenotype, the risk of death was 23% lower in cases published between 1981 and 2000 (hazard ratio, HR, 0.77; 95% CI 0.52–1.14) and 49% lower in those published after 2000 (HR 0.51; 95% CI 0.35–0.73) than in those before 1980 [25]. The authors stress that the data come from case reports and small series, biased towards severe phenotypes and towards countries with greater diagnostic resources, and do not replace a prospective natural history study [25].

3.3 Intermediate phenotype II/III alpha/beta

In the Cathey series, 7 of the 61 probands had a non-homogeneous intermediate phenotype; 3 of those 7 and 5 members of a family described by David-Vizcarra had a homogeneous intermediate phenotype correlated with c.10A>C p.(Lys4Gln) in trans with a frameshift variant [1, 24]. These patients are born with normal anthropometry; the face is sometimes coarse and flat; contractures and mild dorsolumbar kyphoscoliosis appear in infancy; growth slows at the end of the second year and ceases with a maximum height below 90 cm, coinciding with dysostosis multiplex [1]. The initial diagnosis is ML II before 3 years, later changed to ML III or ML II/III, because all walked without support at around 3 years, speech is almost normal in some and always better than in ML II, intellectual disability is mild and social interaction almost normal [1, 4]. Life expectancy approaches that of ML III; respiratory problems are mild and rare, and progressive valvular stiffening with regurgitation, hypertrophy and myocardial failure is the most common cause of death [1]. In the anaesthesia series of Ammer and colleagues, a difficult airway was anticipated in 80% of anaesthetics in patients with the intermediate form [94].

3.4 Mucolipidosis III alpha/beta

3.4.1 Onset and presentation

ML III alpha/beta becomes evident at around 3 years, with slowing of linear growth and joint stiffness and pain initially in the shoulders, hips and fingers [1]. The usual presentation between 2 and 4 years is articular, with pain that may wake the child, and early, progressive stiffness often leads to a presumptive diagnosis of juvenile idiopathic arthritis (JIA), even though the erythrocyte sedimentation rate is normal [40]. Many patients are evaluated for rheumatological disease before diagnosis [1]. In the Turkish paediatric metabolic clinic, ML III accounted for 6 of 192 children referred for joint stiffness or pain; median onset was at 2.75 years, finger joint stiffness was the first sign in all, 5 of 6 had a previous incorrect diagnosis and the median diagnostic delay was 6.5 years (3–24) [87]. In the review of 843 cases, the median age at diagnosis of ML III was 9.0 years [25]. The spectrum includes very mild forms: two siblings with isolated hip involvement, minimal spinal changes and subtle corneal opacities on slit-lamp examination, with markedly increased serum hydrolases (Freisinger and colleagues, 1992), and a 13-year-old girl whose only complaint was hand stiffness (Tylki-Szymanska and colleagues, 2002) [16].

3.4.2 Osteoarticular system

Progressive stiffness of large and small joints is the cardinal sign; limitation of the shoulders, mainly of soft tissues, is among the earliest [1]. Gait is slow, inefficient and often waddling, running is markedly impaired and flexion contractures of the hips and knees give the characteristic crouched standing posture, described as “jockey-like” [1, 40]. Severe secondary osteoarthritis of the hips, with destruction of the proximal femoral epiphyses, and stiffening of soft tissues lead many patients to a wheelchair before or in early adulthood [1]. Kyphoscoliosis appears in early or mid-childhood, with little response to orthoses because of osteopenia and dysplasia, and compromises breathing and walking [1]. The wrists and ankles are less affected; Dupuytren-type palmar contractures begin in late childhood and the clawing becomes moderate to severe; the hands and fingers have almost normal length, unlike in ML II [1]. Bilateral carpal tunnel syndrome is one of the earliest and most characteristic manifestations and is almost universal [1, 31]. Osteoporosis affects the whole skeleton and bone pain is the most distressing symptom, even in patients with limited mobility; painful osteolytic lesions appear in non-ambulant patients [1]. Pain due to osteoporosis worsens in adolescence [1]. Hypoplasia of the odontoid process with atlantoaxial dislocation has been described, and atlantoaxial instability may require cervical arthrodesis [5]. There are very mild cases with isolated hip dysplasia and vertebral changes, identified in late childhood [5, 40].

In the 13 adults of Oussoren and colleagues, all had dysostosis multiplex and progressive secondary osteoarthritis with cartilage destruction and bone lesions in multiple joints; all had undergone several orthopaedic operations since the second or third decade, total hip replacement being the most frequent (61%); carpal tunnel syndrome was present in 12 (92%) and 8 (61%) had undergone surgical release [27].

3.4.3 Growth and appearance

Growth is normal at birth; velocity decreases in late infancy and the deficiency is evident after 3 years; height is often below the 3rd centile in childhood and adult stature well below the family target, without profound short stature [1]. Bone Dysplasias gives an adult height of 144 cm in a Japanese man and normal growth in mild cases [5]. Facial coarsening is gradual and most apparent in profile (full cheeks, depressed nasal bridge, prominent mouth), with mild gingival hypertrophy that does not interfere with eruption; there is no true macrocephaly and the neck is short [1]. Skin thickening is inconstant and mild [1]. The Atlas also describes hirsutism, synophrys and dental crowding with malocclusion [40].

3.4.4 Heart and lung

Mitral and aortic thickening and regurgitation are common from late childhood; left and/or right ventricular hypertrophy is frequent on echocardiography in older patients; PH is possible in some and insufficiently documented; progressive valvular disease may require valve replacement and is a major contributor to adult morbidity [1]. Aortic regurgitation with a diastolic murmur may appear at the end of the first decade and symptomatic heart failure is rare [40]. From late childhood, bronchitis and bronchopneumonia are the most consistent respiratory complications; adults develop restrictive lung disease due to chest stiffness, slowly progressive bronchial sclerosis and interstitial thickening [1]. Death in early or mid-adulthood is often cardiopulmonary, even without pneumonia [1].

3.4.5 Cardiac-predominant forms

Some people with biallelic GNPTAB variants present in adulthood with isolated or predominant cardiac disease. The 47-year-old woman described by Steet and colleagues had dilated cardiomyopathy and mild sensory neuropathy, without skeletal or connective tissue changes, with markedly increased serum lysosomal enzymes [57]. In 2026 three reports were published: a 38-year-old man with chest pain and palpitations, left ventricular ejection fraction of 34% and almost circumferential “ring-like” subepicardial late enhancement on cardiac MRI, who met the Padua criteria for arrhythmogenic left ventricular cardiomyopathy, had a negative 168-gene cardiac panel and was diagnosed by genome sequencing with c.771G>A and c.2502del in GNPTAB, with normal urinary and blood lysosomal biomarkers and enzyme activity, an unremarkable skeleton, normal cognition, subtly coarse facial features, mild corneal opacity and cataracts [58]; a 39-year-old woman with precordial pain, slightly elevated troponin and endomyocardial biopsy with vacuolar degeneration and myelin bodies, initially suspected of Fabry disease, with negative GLA sequencing and globotriaosylsphingosine (lyso-Gb3) and an exome with c.2715+1G>A and the variant of uncertain significance c.1702T>G p.(Phe568Val) in GNPTAB, confirmed by elevated plasma lysosomal enzymes, without skeletal or facial changes [10, 59]; and a family with cardiomyopathy and sudden cardiac death as the presentation of ML III with compound heterozygous variants [95]. The authors of the first report propose including GNPTAB in cardiomyopathy panels [58].

3.4.6 Neurodevelopment, senses and other organs

Neurodevelopment is the most variable feature, from normal to mild delay in motor milestones; language develops at the expected age, without stuttering; the intelligence quotient (IQ) is often normal, but about half of children need educational support, often because of physical limitations [1]. Bone Dysplasias indicates mild to moderate intellectual disability in about 50% and the Atlas a frequent IQ of 70–90; among Oussoren’s adults, 31% had mild cognitive impairment [5, 27, 40]. Otitis is more frequent and some patients have conductive deafness [1]. The corneas appear clear on inspection and the opacity is visually insignificant on slit-lamp examination; hypermetropic astigmatism, mild retinopathy and optic nerve changes are described [1, 5, 40]. The abdomen is prominent when standing, with diastasis recti and small umbilical hernias, usually without organomegaly; the absence of organomegaly is a differential clue with respect to ML II [1]. Renal involvement has been described in two patients [5]. In a 39-year-old man a giant connective tissue naevus on the back was described, formed by coarse collagen fibres of types I, III and VI, with moderate reduction of elastic fibres [16].

3.5 Mucolipidosis III gamma

ML III gamma is a slowly progressive inborn error, mainly skeletal, articular and of connective tissue, with onset in early childhood (median 3.0 years) [7]. The initial manifestation in most patients is finger stiffness, as early as 18 months; in the Tüysüz series, 4 patients presented later, between 6 and 16 years, with knee pain or waddling gait, and early onset with severe stiffness was associated with faster progression in the hands, shoulders and hips [7, 26]. Flexion clawing of the fingers is moderate to severe and worsens; shoulder limitation is common and early and genu valgum is present early in all [7]. The hip is usually affected in late adolescence, later than in ML III alpha/beta, and progresses over years to destruction of the proximal femoral epiphyses, so that total hip arthroplasty is necessary in most [7]. Carpal tunnel syndrome is present in most and becomes clinically significant in the second or third decade [7]. There is scoliosis and hyperlordosis, and atlantoaxial instability requiring surgery has been described in one patient [7]. Chronic pain syndrome (hips, knees, legs and sometimes hands and shoulders) is common, of unknown aetiology and greatly affects quality of life; in some patients there is spinal cord compression due to canal stenosis and vertebral osteoarthritis; osteopenia on dual-energy X-ray absorptiometry (DXA) is common [7]. Nampoothiri and colleagues identified forearm hypertrophy and restricted supination as clinical clues [28].

Birth weight and length are normal; growth slows gradually and height is often below the 10th centile; in the most severely affected patients of the Turkish series, final height was below the 3rd centile [7, 26]. Facial coarsening is more gradual than in ML III alpha/beta, and some patients have limited mouth opening, condylar defects and impacted teeth [7, 90]. The corneas are clear on inspection, with opacities without impairment on slit-lamp examination in some, and there is one case of late-onset retinitis pigmentosa [7]. Pulmonary involvement is generally absent, with mild to moderate restrictive disease in adults owing to spinal and rib changes [7, 27]. Valvular thickening, mitral prolapse and mitral and aortic regurgitation are common from late childhood, but cardiac function is normal in most adults and valve replacement is rarely necessary [7]. Motor milestones may be delayed, language and learning are age-appropriate and less than 5% have mild cognitive impairment, whose relationship to ML III gamma is yet to be established [7]. The skin becomes slightly thickened; in patients of Moroccan ancestry scleroderma-like changes have been described [7, 96]. There is no hepatomegaly or splenomegaly [7]. In Brazil, ML III alpha/beta showed more somatic anomalies and early delay than ML III gamma [86].

3.6 Natural history and treated disease

There is no disease-modifying treatment, so the data above describe the natural history under supportive care. Patients with ML II who underwent HSCT continued to have skeletal progression and, in the international series, 27% survival; the results are described in section 7 [30, 97]. In patients with ML III treated with bisphosphonates, the reported improvement is symptomatic (pain and mobility), with no demonstration of a change in the articular natural history [1, 32].

4. Clinical Recognition and Warning Signs

4.1 When to suspect, by age and setting

4.2 Clinical clues that separate the entities of the class

Marked gingival hypertrophy in the first weeks of life, thick skin, periosteal cloaking and shoulder restriction despite hypotonia indicate ML II [1, 40]. Progressive stiffness with absent or mild organomegaly and normal or near-normal cognition indicates ML III; the distinction between ML III alpha/beta and gamma is neither clinical nor biochemical and requires molecular testing, although early genu valgum and later hip involvement point to the gamma form [1, 7]. Independent walking at around 3 years in a child with the physical appearance of ML II suggests the intermediate phenotype [1].

4.3 Warning signs and emergency situations

Any lower respiratory infection in a child with ML II may precipitate respiratory failure, and the persistent assisted ventilation of the longest survivors always began during treatment of an acute infection [1]. The need for sedation or anaesthesia is a risk situation in all forms: a difficult airway was anticipated in 100% of anaesthetics in patients with ML II and a failed intubation attempt ended in death in the Manchester series [68, 94]. Neck pain, pyramidal signs or limb weakness require exclusion of spinal cord compression and atlantoaxial instability [5, 7, 102]. Exertional dyspnoea, syncope or palpitations in a patient with ML III require assessment for valvular disease, ventricular dysfunction, PH and arrhythmia, and subclinical heart failure may manifest during an anaesthetic [1, 58]. Sudden bone pain in a patient treated with bisphosphonates raises the possibility of atypical femoral fracture, and jaw pain that of osteonecrosis of the jaw [1, 103].

4.4 Newborn screening

The sources consulted do not describe the inclusion of ML II or ML III in newborn screening programmes in the USA, Europe, Portugal or Brazil. A multiplex tandem mass spectrometry assay measuring hydrolase activities in dried blood spots has been developed for screening and diagnosis, and hydrolases may be markedly elevated in the dried blood spots of patients with ML III alpha/beta and ML III gamma [7, 29]. ML II is, in most cases, clinically evident at birth, and the ML III forms only become manifest after the first year [1].

5. Diagnosis: Criteria, Tests and Algorithms

5.1 Diagnostic criteria

There are no published consensus clinical diagnostic criteria for any of the entities of the class [1, 7]. The diagnosis is established in a proband with suggestive clinical, radiological and biochemical findings by the identification of biallelic pathogenic or likely pathogenic (P/LP) variants in GNPTAB or GNPTG, the P and LP classes being equivalent for clinical decision-making according to the standards of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) [1, 7]. Identification of biallelic variants of uncertain significance (VUS), or of one P/LP variant with one VUS, neither establishes nor excludes the diagnosis [1]. GlcNAc-1-phosphotransferase deficiency in fibroblasts confirms the biochemical diagnosis [2, 7]. A family history compatible with autosomal recessive transmission (affected siblings, consanguinity) supports the suspicion, but its absence does not exclude it [1].

5.2 Algorithm

5.3 Biochemistry and biomarkers

Elevation of several hydrolases in plasma or serum is the central biochemical finding. The GNPTG GeneReviews indicates activities up to 10 times higher in serum, dried blood and other fluids, including fibroblast or amniocyte culture medium, and highlights as the enzymes of greatest interest in the differential beta-D-hexosaminidase (EC 3.2.1.52), beta-D-glucuronidase (EC 3.2.1.31), beta-D-galactosidase (EC 3.2.1.23), alpha-D-mannosidase (EC 3.2.1.24), alpha-L-iduronidase (EC 3.2.1.76) and iduronate-2-sulfatase (EC 3.1.6.13) [7]. Smith's Recognizable Patterns indicates activities 5 to 20 times higher in plasma and the Atlas up to 100 times for some enzymes [40, 44]. In the Italian case of ML III gamma described in section 10, total plasma beta-hexosaminidase was 7868 nmol/ml/h (reference value 883 ± 252) and alpha-L-fucosidase 1177 nmol/ml/h (595 ± 193) [14]. In a Brazilian series of five patients, alpha-iduronidase, alpha-mannosidase, beta-glucuronidase and hexosaminidases A and total were increased in plasma, with normal GAGs [108]. Reference values depend on the laboratory, so the report should state the unit (nmol/ml/h or equivalent) and the local reference interval.

Biochemistry has four practical limitations: it does not distinguish ML III alpha/beta from ML III gamma; it does not identify heterozygotes; ML III gamma is not diagnosed by hydrolase activity in leucocytes; and leucocyte activity in ML II is described as normal by some sources and as increased by others, so the diagnosis should not be based on it [1, 7, 63]. Intracellular hydrolase activity in cultured fibroblasts is low compared with controls and supports the diagnosis; acid beta-glucosidase remains normal [7, 40]. GlcNAc-1-phosphotransferase activity measured in fibroblasts is below 0.02 pmol/mg/h in ML II and in some ML III, against 0.67–1.46 in controls [47]; with alpha-methylmannoside as acceptor, activity was less than 0.4 to 2.0 pmol/mg/h in ML II and 2.9 to 39.4 in ML III, and is normal in the gamma variant [49]. In dried blood spots, a multiplex tandem mass spectrometry assay measures several activities and has been proposed for screening and diagnosis; the 2025 Turkish series used beta-hexosaminidase A and A+B, alpha-D-mannosidase and alpha-L-fucosidase activities in dried blood as minimally invasive screening, concordant with the molecular diagnosis [29, 63].

Elevation of several hydrolases in plasma is not exclusive to ML II/III. Hereditary fructose intolerance, in an infant fed a fructose-containing formula and with liver failure, reproduced the biochemical profile of mucolipidosis; a normal plasma alpha-mannosidase with the other hydrolases elevated should raise this possibility, and the review by Ferreira and colleagues lists other secondary causes of elevated serum lysosomal enzymes [106]. The reverse also occurs: three children with ML III homozygous for the common low-activity alpha-L-fucosidase variant had normal plasma fucosidase, although it is usually elevated in ML III, so a single normal enzyme does not exclude the diagnosis [16].

Urinary analyses give divergent results between sources. Bone Dysplasias and Smith's Recognizable Patterns describe increased oligosaccharide excretion usable as screening, whereas Inborn Metabolic Diseases warns that patients with ML II and ML III are often missed on urinary tests and that definitive diagnosis rests on plasma elevation of several enzymes [5, 31, 44]. Traditional thin-layer chromatography may miss cases; ultra-high-performance liquid chromatography platforms with high-resolution mass spectrometry (UHPLC/HRAM-MS) identified ML II/III in patient samples, with the pitfall of occasional “neonatal” profiles of probably nutritional origin in infants [31, 109]. Urinary GAG screening is typically normal or only slightly elevated; urinary keratan sulphate may be elevated [1]. In the multiplex assay of Langereis and colleagues by liquid chromatography–tandem mass spectrometry (LC-MS/MS), with 100% sensitivity and 92% specificity in 100 patients with MPS, the three patients with ML II/III had elevated GAGs by LC-MS/MS, but the colorimetric dimethylmethylene blue (DMB) test was normal in two of them [105]. Urinary excretion of acid mucopolysaccharides is normal in ML II, and this absence of mucopolysacchariduria was the finding that led to the original description [2, 43].

5.4 Radiology

The dysostosis multiplex of ML II is severe before 2 years and that of ML III is mild and apparent between 5 and 10 years [1]. Bone Dysplasias considers that Morquio disease, ML II and ML III have distinctive radiological features within the group [104].

In the Brazilian series of five patients with ML II/III, the most constant findings were anterior beaking of the vertebral bodies, shallow acetabula and epiphyseal dysplasia (5/5), followed by platyspondyly, pelvic dysplasia and reduced mineralisation (4/5) [108]. In ML III, MRI of the hips and knees before considering arthroplasty and baseline DXA in children over 5 years and in adults are part of the assessment recommended by the GNPTG GeneReviews [7].

5.5 Cytology and pathology

Cultured fibroblasts show dense inclusions on phase microscopy, and there are inclusions or vacuolation in biopsied cornea, bone marrow and lymphocytes [40]. Vacuolated lymphocytes, combined with coarse face, bone changes and stiffness, suggest mucolipidoses or MPS [89]. The placenta may provide the diagnosis: vacuolation of the syncytiotrophoblast with lipid vacuoles stained by Oil Red O in a newborn admitted for respiratory distress led to ultra-rapid genome sequencing and identification of compound heterozygous variants in GNPTAB, concordant with the exome of placental tissue [99]; other reports describe placental vacuolation, including a dichorionic twin pregnancy with only one fetus and one placental disc affected [100, 112]. Endomyocardial biopsy with vacuolar degeneration and myelin bodies on electron microscopy may mimic Fabry disease [59].

5.6 Molecular testing and its limitations

These percentages are proportions of known variants detectable by each method and are not equivalent to the clinical sensitivity of the test. The GNPTAB GeneReviews proposes two routes: GNPTAB sequencing followed by deletion and duplication analysis if zero or one variant is found, or a multigene panel with GNPTAB and genes of the differential; and exome or genome sequencing when the phenotype is atypical, in line with the ACMG and American Academy of Pediatrics recommendations for developmental delay, intellectual disability and multiple congenital anomalies [1]. For GNPTG, the recommended route is a mucolipidosis or lysosomal disease panel that includes the gene, or the exome or genome [7]. Cardiomyopathy panels may not include GNPTAB: in the 2026 case, a 168-gene panel was negative and only trio genome sequencing identified the variants [58].

The limitations to consider are as follows. Exon sequencing does not detect deep intronic variants, such as GNPTG IVS7-10G>A (NM_032520.5:c.527-10G>A) or the intronic deletions described by Persichetti, if these lie outside the region analysed [80, 81]. Splice variants outside the canonical positions and synonymous variants, such as c.771G>A in the last nucleotide of exon 7 of GNPTAB, may be filtered out as benign without RNA studies [57, 58]. Apparent homozygosity may correspond to a deletion of one or several exons in trans not detected by sequencing, or to uniparental isodisomy, and requires parental testing [1, 7]. When only one parent is a carrier, with confirmed parentage, the other variant may have arisen de novo in the proband or through parental germline mosaicism; this situation has been described in a Brazilian family with ML III gamma [7, 107]. In the Portuguese and Brazilian series, the effect of splice variants was confirmed by cDNA analysis, which shows the usefulness of RNA studies when the variant is of doubtful interpretation [71, 76]. A negative molecular result, with concordant phenotype and biochemistry, does not exclude the disease and justifies RNA studies, genome sequencing or measurement of phosphotransferase activity.

The molecular report should state the gene, the reference transcript with version (NM_024312.5 for GNPTAB; NM_032520.5 for GNPTG), the HGVS description of the variant at cDNA and protein level, the zygosity, the phase when known, the ACMG/AMP class and the criteria applied [1, 10]. Historical notations, such as c.3503_3504delTC or 500insC, should be converted to the current form (c.3503_3504del, c.499dup) using validation tools, and not by hand [10, 69]. VUS, such as c.1702T>G p.(Phe568Val) in the cardiac-predominant Chinese case, require segregation, biochemistry and, if possible, functional studies [10, 53, 59].

5.7 Assessment after diagnosis

5.8 Prenatal and preimplantation diagnosis

Once the familial variants have been identified, molecular prenatal diagnosis on chorionic villi or amniocytes and preimplantation genetic testing are possible and are the preferred method [1, 44]. In Korea, two molecular prenatal diagnoses on chorionic villi were successful [73]. Without known variants, biochemical prenatal diagnosis in the first trimester is possible by the activity of several lysosomal enzymes in cell-free amniotic fluid or by the phosphotransferase assay on chorionic villi, and electron microscopy shows vacuolation of villous cells [2, 40, 44]; the usefulness of measuring GlcNAc-1-phosphotransferase in first-trimester chorionic villi was demonstrated by Ben-Yoseph and colleagues in 1988 [15]. Ultrasound does not show changes in all affected fetuses [44]. In non-immune hydrops fetalis, the Oxford Desk Reference recommends considering biochemical analysis of amniotic fluid and cultured amniocytes for lysosomal diseases, in addition to a panel, exome or genome; in an Indian series, 7 of 33 cases of non-immune hydrops fetalis (21%) had a lysosomal disease, one of them mucolipidosis, and the proportion was 4 of 11 (36%) in cases of recurrent hydrops [98, 113].

6. Differential Diagnosis

6.1 Principles

The differential depends on age and setting of presentation: in the newborn and infant, the other lysosomal diseases with severe dysostosis multiplex and neonatal hyperparathyroidism; in the child with joint stiffness, the attenuated MPS, skeletal dysplasias and inflammatory arthropathies; in the adult with cardiomyopathy, the other storage and genetic cardiomyopathies. The most frequent misdiagnoses of ML III alpha/beta are attenuated MPS I and JIA [1]. The test that separates ML II/III from almost all other lysosomal diseases is the simultaneous elevation of several hydrolases in plasma, which is absent in the MPS, the oligosaccharidoses and the dysplasias [1, 7].

6.2 Comparative table

6.3 Main differentials

MPS. Overlap with attenuated MPS I is the most frequent and, in the series of adults with ML III, 31% had initially been diagnosed as MPS [1, 27]. In ML II, neonatal changes are more severe than in the MPS, macrocrania is rare, corneal opacities are mild and urinary mucopolysaccharides are not elevated [2]. In ML III, the vertebral changes are different, the tubular bones of the hands and feet are less affected and the fingers are relatively long and thin, whereas in the MPS and ML II the hands are short with contracted fingers [5]. The colorimetric urinary GAG test may be normal both in ML and in some attenuated MPS; therefore, the plasma enzyme panel should be part of the work-up of a suspected MPS with normal GAGs [1, 105].

GM1 and galactosialidosis. In the infant, GM1 gangliosidosis type I and galactosialidosis are morphologically very similar to ML II and can only be excluded by enzyme analysis; a cherry-red spot, rapid neurological degeneration and hepatomegaly point to GM1, and contractures are more severe in ML II [2].

Rheumatological diseases. Finger stiffness without synovitis, claw hand, normal erythrocyte sedimentation rate, forearm hypertrophy and limited supination are the clues to ML III, and the correct diagnosis avoids unwarranted immunosuppressants and disease-modifying antirheumatic drugs [28, 40, 87]. In the Italian case described in section 10, the child was followed at a paediatric rheumatology centre, with negative tests for Farber disease and urinary GAGs, before the plasma enzyme panel pointed to the diagnosis [14]. The familial scleroderma-like picture described in Morocco with GNPTG variants shows another form of confusion with rheumatic disease [96, 117].

Skeletal dysplasias. Spondyloepiphyseal dysplasias, progressive pseudorheumatoid dysplasia and multiple epiphyseal dysplasias cause stiffness and osteoarthritis without typical dysostosis multiplex (rounded hook-shaped vertebrae, hypoplastic and oblique lower ilia, short metacarpals with proximal pointing, “bullet-shaped” phalanges) and without coarse features [2, 5].

Neonatal hyperparathyroidism. Idiopathic neonatal hyperparathyroidism may cause periosteal cloaking in the newborn, but the lesions are generally milder, without dysostosis multiplex and with normal lysosomal enzymes; I-cell disease presents with facial dysmorphism and may show stippled epiphyses in the tarsal and lumbosacral regions [2, 101]. The primary severe form results from loss-of-function variants in CASR, usually biallelic (the parents may have familial hypocalciuric hypercalcaemia), but also from heterozygous de novo variants; it manifests in the first 6 months with severe hypercalcaemia, bone demineralisation and growth retardation, and parathyroidectomy is reserved for infants in whom intensive medical therapy does not stabilise the disease [116].

Mucolipidosis IV. ML IV presents at the end of the first year with severe psychomotor delay and progressive visual impairment due to corneal opacity and retinal degeneration, with achlorhydria and elevated plasma gastrin in practically all patients; there is no dysostosis multiplex or elevation of plasma hydrolases, and the defect lies in the cation channel mucolipin-1, not in the M6P pathway [41]. The carrier frequency of the two common Ashkenazi variants is 1:100 to 1:127 [41].

Cardiomyopathies. In the adult with dilated or arrhythmogenic cardiomyopathy, with a biopsy showing lysosomal storage and negative GLA, or with a negative cardiomyopathy panel, cardiac-predominant ML III alpha/beta is a hypothesis to consider, even with a normal skeleton; plasma hydrolases were elevated in one case and normal in another [57–59].

7. Treatments and Risk Management Strategies

7.1 General principles

There are no published clinical practice guidelines for GNPTAB-related disorders or for ML III gamma; the recommendations of the two GeneReviews chapters are based on the authors’ personal experience, and the search for consensus documents carried out for this article found only the OrphanAnesthesia anaesthesia recommendation [1, 7, 92]. There is no authorised disease-modifying therapy. Treatment is supportive, multidisciplinary and delivered in specialised centres, with palliative care integrated early in ML II; ear, nose and throat (ENT) surgery, orthopaedics and neurosurgery may be needed, and anaesthetic considerations are generally less severe than in the MPS [31]. In ML II, the aim is comfort, respiratory function, feeding and interaction; in ML III, preservation of mobility, pain control, cardiac surveillance and orthopaedic surgery at the appropriate time [1, 7].

7.2 Supportive treatment by manifestation

7.3 Rehabilitation and procedures to avoid

The joint pain of ML III is exacerbated by strenuous exercise and vigorous physiotherapy, and GeneReviews recommends low-impact conservative management [1]. Vigorous stretching is ineffective and painful and may damage the joint capsule and tendons; in ML II, aggressive therapy may cause soft-tissue calcification, and the physiotherapy that benefits other children cannot be unequivocally recommended [1, 7]. In forms mistaken for rheumatic disease, the correct diagnosis avoids unwarranted immunosuppressants and disease-modifying antirheumatic drugs [28]. Suspicion of craniosynostosis, frequent because of metopic prominence, should not lead to cranial surgery without confirmation, because in some cases it has prompted inappropriate surgery [1].

7.4 Orthopaedic surgery

Bilateral hip arthroplasty has been successful in older adolescents and adults with milder ML III, as has knee arthroplasty, with intensive rehabilitation after surgery; the timing of arthroplasties is discussed between the surgeon and the patient [1, 7, 118]. The Atlas of Inherited Metabolic Diseases recommends postponing hip surgery until after puberty [40]. The review by Oussoren and colleagues on hip disease in the MPS and ML discusses the mechanisms (abnormal pelvic and femoral development, biomechanical forces, inflammation), interventions and prospects [31, 119]. Spinal deformities are treated by spinal surgeons, with surgery for severe deformities, with or without spinal cord compression, and atlantoaxial instability may require cervical arthrodesis [5, 7]. Patients should be operated on in centres with experience in skeletal dysplasias and metabolic bone disease, because anaesthesia adds its own risk (7.8) [1].

7.5 Bisphosphonates, vitamin D and calcium

7.5.1 Rationale and evidence

Bisphosphonates decrease osteoclastic activity and have been proposed for the bone pain of ML III, with variable success [1, 31]. The rationale is the increased bone resorption in patients with GNPTAB variants, demonstrated on biopsies and urinary markers, which is not seen in ML III gamma; the authors of that study therefore recommend classifying ML III also by gene before considering antiresorptive therapy [56]. The main human evidence is the study by Robinson and colleagues (2002): two siblings with ML III in New Zealand, with increased remodelling markers and a transiliac biopsy showing trabecular osteopenia and marked subperiosteal resorption, received monthly intravenous pamidronate for 1 year, well tolerated, with marked reduction in bone pain and improved mobility, despite incomplete suppression of resorption on biochemistry, radiography and histology [32, 40]. GeneReviews adds that some patients, but not all, report less pain and greater mobility and that densitometry improves [1]. For ML III gamma, GeneReviews advises considering oral or intravenous bisphosphonates if there is significant skeletal disease and a DXA Z-score of −2.5 or below, based on the Tüysüz series and the personal observation of one of the authors [7, 26]. Bone Dysplasias recommends 25-hydroxycolecalciferol and bisphosphonates to accelerate healing of the early osteopenia of ML II, without stating doses; in the Irish series, neonatal hyperparathyroidism resolved in all patients, whether or not treated with high-dose vitamin D [2, 22]. In a murine model of ML II (2026), zoledronic acid improved bone density, articular cartilage, motor deficit and hepatomegaly and partially restored M6P-dependent hydrolases, without human data [120].

7.5.2 Doses

There is no published bisphosphonate dose specific to ML II or ML III: the abstract of the Robinson study does not state the pamidronate dose and the full text was not accessible [32]. The following doses are extrapolated from paediatric osteogenesis imperfecta (OI), a condition treated with the same drugs and with published protocols, and from the paediatric trial described in the Aclasta® EPAR; all correspond to off-label use in ML. The reference is the 2018 Australasian Paediatric Endocrine Group (APEG) consensus on bisphosphonates in children and adolescents (Simm and colleagues), which grades recommendations using the GRADE system (1 = strong, 2 = weak; quality of evidence from very low to high) and is the only paediatric guideline consulted that gives doses; centre protocols and labels serve as comparison [121]. The consensus does not address ML; its indications for secondary osteoporosis and for bone pain are the closest to the situation of patients with ML III [121].

In secondary osteoporosis, the consensus proposes considering intravenous bisphosphonate in children with vertebral fracture(s) and/or low bone mineral density with two or more long-bone fractures (recommendation 1, low evidence), only after optimising vitamin D, calcium intake, mobility and pubertal induction when necessary; prophylactic treatment of a low Z-score without fracture is not recommended and, in cerebral palsy, is rarely considered outside severe pain presumed to be of bone origin [121]. The duration in secondary osteoporosis is not defined: the authors propose annual DXA and treatment beyond 2 years only if fractures or bone pain persist [121]. Outside bone fragility, the consensus accepts intravenous bisphosphonates for the pain of fibrous dysplasia (24 months of pamidronate or zoledronic acid; recommendation 1) and for the pain of avascular necrosis (recommendation 2, very low evidence), without demonstrated effect on the structural course of the lesions [121]. The indication in ML III for bone pain thus rests on an extrapolation from conditions in which analgesic benefit is the main outcome [32, 121].

7.5.3 Monitoring, adverse effects and contraindications

Before the first cycle, the GOSH protocol requires a lateral radiograph of the whole spine, lumbar DXA (without normative values below 5 years, but useful as the patient’s own control above 10 kg), renal ultrasound for nephrocalcinosis and measurement of calcium, phosphate, urea and electrolytes, alkaline phosphatase, albumin, intact parathyroid hormone and 25-hydroxyvitamin D; before day 1 of each cycle the biochemistry is repeated and, on days 2 and 3, calcium, phosphate and albumin; after 1 year and then annually the baseline tests are repeated, with DXA on the same machine [122]. The acute phase reaction (fever, generalised aches, vomiting) occurs in most children after the first infusion, lasts 24 to 48 hours and is treated symptomatically with paracetamol; in newborns it may include respiratory distress, which is relevant in ML II because of pre-existing respiratory disease [122]. Hypocalcaemia is rarely symptomatic; if it occurs, calcium supplements are doubled for 5 days [122]. In the paediatric trial of zoledronic acid in OI, the most frequent effects were pyrexia (61%), arthralgia (26%), hypocalcaemia (22%) and headache (22%), mainly in the 3 days after the first infusion [127]. The labels warn of hypocalcaemia (to be corrected beforehand), renal impairment, osteonecrosis of the jaw (prior oral examination), atypical femoral fractures and severe musculoskeletal pain [127]. GeneReviews states that atypical femoral fractures and osteonecrosis of the jaw are rare complications, citing a patient with both on oral bisphosphonate [1, 103]. The GOSH protocol recommends not giving pamidronate in the 2 to 4 weeks before orthopaedic surgery and withholding it until there is radiological union, because of the risk of delayed union; this point is relevant in ML III, in which osteotomies and arthroplasties are frequent [122]. The protocol’s contraindications are hypersensitivity, renal impairment with a glomerular filtration rate below 30 ml/min/1.73 m², pregnancy and vitamin D deficiency; girls of childbearing age should be asked about pregnancy and, if in doubt, tested before the infusion [122]. The US pamidronate label contraindicates it in hypersensitivity to bisphosphonates [125].

The 2018 APEG consensus adds safety and monitoring recommendations [121]. Before the first dose it requires a full blood count, urea and electrolytes, liver function tests, 25-hydroxyvitamin D, parathyroid hormone, calcium, magnesium and phosphate, dental review, DXA and a lateral radiograph of the thoracolumbar spine (or vertebral morphological assessment by DXA); DXA is repeated annually, but reference values are lacking below 3 years and there are technical limitations with significant contractures or orthopaedic hardware, a frequent situation in ML III [121]. Vitamin D should be above 50 nmol/L before the first infusion and calcium intake should be maintained afterwards; calcitriol for 3 days after the first dose may attenuate hypocalcaemia; up to 80% of patients have an acute phase reaction in the 24–48 hours after the first infusion, treated with paracetamol, antiemetics and fluids, and a reduced first dose lessens these effects (recommendation 1, low evidence) [121]. Dental review is done before the first dose, with invasive dental treatment completed beforehand, and thereafter every 6 to 12 months; there are no reports of osteonecrosis of the jaw in children [121]. “Atypical” femoral fractures in young people on bisphosphonates may be unrelated to the drug and do not in themselves require it to be stopped [121]. In post-menarchal girls, a qualitative pregnancy test is performed before each dose, and pregnancy should be avoided for 12 months after a dose [121]. Oral bisphosphonates should only be used in children who can safely swallow a whole tablet with a glass of water and who do not have gastro-oesophageal reflux, because of the risk of erosive oesophagitis, which in children with ML adds to the postural difficulties already mentioned [121]. Whenever possible, treatment is withheld until there is callus at the site of a fracture or osteotomy (recommendation based on OI), a criterion close to that of the GOSH protocol, which requires radiological union and a 2- to 4-week pause before orthopaedic surgery [121, 122]. The consensus contraindications are pregnancy, renal impairment (with “extreme caution”, without a numerical threshold, whereas the GOSH protocol sets 30 mL/min/1.73 m²), diseases in which inhibition of resorption worsens fragility (hypophosphatasia, sclerotic lesions, high bone mass disorders) and active rickets, which first requires correction of mineral deficiencies [121, 122]. By extension of this last contraindication, the neonatal phase of ML II with hyperparathyroidism and rickets-like changes should be reassessed before any bisphosphonate; the consensus does not specifically address this situation [2, 121].

7.6 Regulatory status

No medicinal product is authorised for ML II or ML III in any jurisdiction consulted, and the European Medicines Agency (EMA) orphan designations report, updated on 7 October 2026, contains no designation for mucolipidosis, I-cell disease or GNPTAB [34]. In the orphan designation database of the Office of Orphan Products Development (OOPD) of the US Food and Drug Administration (FDA), consulted on 27 and 28 October 2026, a search for “mucolipidosis” returns a single record: an adeno-associated virus type 9 vector with the S1S3 gene, which encodes an S1S3 variant of GlcNAc-1-phosphotransferase, designated on 20 October 2020 for the “treatment of mucolipidosis”, with “designated” status and without approval; a search for “I-cell” returned no records [35]. A news item from February 2021 attributes this designation to M6P Therapeutics; the fields extracted from the FDA database do not state the sponsor, and the text search of the designation does not exclude records with other wording [35, 129]. Bisphosphonates are therefore used off-label.

Sources for the table: labels and SmPCs consulted [34, 35, 124–130].

7.7 Haematopoietic stem cell transplantation

The rationale for HSCT is cross-correction: donor-derived cells would secrete M6P-bearing enzymes that could be taken up by the patient’s cells. Clinical results in ML II have not confirmed benefit. In the international experience compiled by the Center for International Blood and Marrow Transplant Research (CIBMTR) and by North American centres, 22 children with ML II were transplanted, with engraftment in 95%, but only 6 (27%) were alive at last follow-up; the most common cause of post-transplant death was cardiovascular complications, probably due to disease progression, and survivors had global delay and frequently needed gastrostomy, tracheostomy and mechanical ventilation [30]. Inborn Metabolic Diseases counts 30–40 transplanted patients with ML II/III and classifies HSCT as “not indicated”, in contrast with neuronopathic MPS I (more than 200 patients, standard treatment) [31]. In the review of 843 published cases, 32 children with ML II received stem cell transplantation (marrow, peripheral blood or umbilical cord); 14 died, with a median age at death of 2.9 years, similar to that of non-transplanted children (2.7 years), and median survival of 9.0 years (95% CI 4.42–13.58) versus 5.0 years (95% CI 3.91–6.09) did not differ significantly (p = 0.239) [25]. The other potentially life-prolonging procedures described in ML II were tracheostomy (9 of 70), gastrostomy (7), craniosynostosis surgery (7), cervical correction (5) and cardiac interventions (4), and in ML III cardiac surgery — valve replacement, heart transplantation, implantable cardioverter-defibrillator — (7 of 16), cervical decompression or stabilisation (5) and stabilisation of severe kyphosis (2) [25].

Individual reports describe partial effects. A girl transplanted at 19 months with marrow from a human leucocyte antigen (HLA)-identical carrier sibling was, at 7 years, free of respiratory infections, with a shortening fraction of 47% and slowly gaining developmental milestones, but with worsening musculoskeletal deformities [44, 131]. In Hamburg, a girl diagnosed at 3 months and transplanted at 9 months with an HLA 9/10 unrelated donor had a sustained reduction in markers of lysosomal storage and bone metabolism; at 6 years she had normal cardiac function, partial respiratory insufficiency, moderate hepatomegaly and skeletal progression, walked and had a developmental quotient of 36%; she died at 6.6 years after human metapneumovirus pneumonia, and the authors considered the benefit uncertain owing to the lack of comparable natural history data [97]. In China, a child transplanted at 12 months had motor improvement 4 months later, with benefit considered uncertain [132]. In Japan, umbilical cord blood transplantation at 26 months (HLA 3/6) after conditioning with fludarabine, cyclophosphamide, 6 Gy total body irradiation and rabbit antithymocyte globulin reduced plasma enzyme activities by about 20–40% and normalised immunoglobulins A, G2 and G4, but the child had bacterial pneumonias with acute respiratory distress syndrome at 37 and 38 months, with limited clinical effect [133]. The Minnesota review of HSCT in the glycoproteinoses concludes that results vary greatly between diseases [134]. In practice, HSCT in ML II should only be discussed within research or informed shared decision-making about these results.

These sources include no controlled trial, no confidence intervals and no comparison group; the effects described in single cases cannot be compared with each other or with the natural history, which is variable and known mainly from published series. The ClinicalTrials.gov search identified no published negative trials, suspensions or withdrawals of drugs for ML II/III, with the exception of the observational biomarker study BioML, withdrawn without enrolment [33].

7.8 Anaesthesia and the perioperative period

The most frequent operations in ML II/III are adenoidectomy, tonsillectomy, myringotomy with tubes, herniotomy, carpal tunnel release and endoscopic gastrostomy [92]. In 44 anaesthetics in 12 patients (7 ML II, 2 intermediate, 3 ML III) at two centres between 2008 and 2022, there was at least one complication in 27.3% of anaesthetics, 94% of complications occurred in patients with ML II or the intermediate form, 59% arose at induction, respiratory complications accounted for 18% and difficult airway for 14%; a difficult airway had been anticipated in all ML II anaesthetics and in 80% of those of the intermediate form [94]. In Manchester, in 11 anaesthetics in 5 children with ML II/III, fibreoptic intubation or videolaryngoscopy was used in 6 of 7 intubations, 4 of 7 required a change from the initial method, 4 of 5 patients had an abnormal echocardiogram and, of the 3 with cervical radiological assessment, 2 had changes and 1 signs of atlantoaxial instability [102].

The OrphanAnesthesia recommendation (last modified in June 2011, published in 2015) advises, because of the high rate of complications of general anaesthesia in ML II, considering alternatives: carpal tunnel release is possible with ultrasound-guided axillary brachial plexus block, and sedation is decided case by case [92]. Preoperative assessment includes mandibular and cervical stiffness, the short neck, gingival hypertrophy, fragile teeth, thickening of the tongue base, larynx, epiglottis and tracheal wall, enlarged adenoids and tonsils, OSA (with polysomnography if there is a suggestive history) and echocardiography (valvular disease, dilated or hypertrophic cardiomyopathy, PH and, rarely, coronary occlusion) [92]. Airway management follows a difficult airway algorithm, with an experienced team, laryngeal masks of several sizes, a small fibrescope and an endotracheal tube smaller than predicted for age and size; a nasopharyngeal tube should be avoided because of the anteverted nares and the adenoids, with a risk of bleeding; secure venous access is obtained before induction, with intraosseous equipment available [1, 92]. Sedative premedication is relatively contraindicated because of OSA and respiratory insufficiency, and antimuscarinic premedication may be useful; it is recommended to maintain spontaneous ventilation after inhalational induction or with titrated propofol, early CPAP if there is obstruction, a short-acting muscle relaxant only if mask ventilation is effective, fibreoptic intubation through a laryngeal mask if laryngoscopy fails, capnography always, early extubation and postoperative monitoring; day-case surgery is not recommended [92]. GeneReviews adds that surgery should be avoided as far as possible and performed only in a tertiary centre with paediatric anaesthesia and intensive care, that extubation is difficult, that difficulty increases with age in ML III and that subclinical heart failure may become manifest during anaesthesia, justifying a full cardiac assessment before any surgery [1]. In ML III gamma, limited mobility of the jaw and neck, the short neck, atlantoaxial instability and chest restriction require multidisciplinary planning and an available fibrescope [7]. Elective procedures should be performed by an experienced paediatric anaesthetist, in a tertiary centre with paediatric ENT and on-site intensive care, after multidisciplinary planning and risk–benefit analysis [94, 102]. Pamidronate should not be given in the 2 to 4 weeks before orthopaedic surgery [122].

7.9 Prophylaxis of infective endocarditis and divergence between sources

The ML III gamma GeneReviews recommends antibiotic prophylaxis before minor and major surgical procedures, including dental procedures, in patients with valvular dysfunction [7]. Current cardiology guidelines are more restrictive. The 2021 American Heart Association (AHA) scientific statement suggests prophylaxis before dental procedures only in patients with a prosthetic valve or valve repair material, previous endocarditis, certain congenital heart diseases and heart transplant recipients with valvular disease, and emphasises oral health [135]. The 2023 European Society of Cardiology (ESC) guideline recommends prophylaxis in high-risk patients (previous endocarditis, surgical or transcatheter prosthetic valves, valve repair material, untreated cyanotic congenital heart disease or with prosthetic material, ventricular assist devices) undergoing extractions, oral surgery and procedures involving manipulation of the gingival or periapical region, and classifies non-rheumatic degenerative valve disease and congenital valve anomalies as intermediate risk, for which it recommends general preventive measures but not antibiotic prophylaxis [136]. Thus, according to the AHA and the ESC, a patient with ML III and native valvular disease has no indication for antibiotic prophylaxis, but has an indication for oral and skin hygiene measures; after valve replacement or repair, there is an indication for prophylaxis before at-risk dental procedures [135, 136]. This article follows the cardiology guidelines and records the broader GeneReviews recommendation as a divergence.

The doses are single doses, 30 to 60 minutes before the procedure; the AHA accepts administration up to 2 hours afterwards if it has been forgotten; cephalosporins should not be used in people with a history of anaphylaxis, angioedema or urticaria with penicillin or ampicillin; clindamycin is no longer recommended [135, 136].

7.10 Pregnancy, contraception and lactation

The literature search carried out for this article (PubMed, October 2026) found no reports of pregnancy in women with ML II or ML III, and this absence does not demonstrate that none exist. Women with ML III reach adulthood and their offspring are obligate carriers [1, 7]. Restrictive lung disease, valvular disease, kyphoscoliosis, pelvic dysplasia and a difficult airway are factors that justify preconception counselling and obstetric and anaesthetic planning in an experienced centre, although the sources do not describe obstetric anaesthesia in these patients [1, 92]. Bisphosphonates are contraindicated in pregnancy in the GOSH protocol, and pregnancy should be excluded before each cycle in girls of childbearing age [122].

7.11 Transition to adulthood and palliative care

In ML III, transition coincides with the period in which hip disease, bone pain, osteoporosis and valvular disease become decisive and in which many patients become wheelchair-dependent; among the adults of Oussoren and colleagues, 46% needed help in activities of daily living or a wheelchair, and all had undergone surgery several times [1, 27]. Continuity with adult internal medicine or metabolic medicine, orthopaedics, pain medicine, cardiology and respiratory medicine replaces paediatric follow-up. In ML II, palliative care, respite for the caregiver and home nursing are part of the plan from diagnosis [1, 31].

7.12 Guidelines, consensus documents and divergences

Sources for the table: [1, 7, 92, 121, 122, 135, 136].

7.13 Registered clinical studies

Source: search of the ClinicalTrials.gov application programming interface on 27 October 2026, with no interventional trial specific to ML II/III; a phase 1 gene therapy trial with an adeno-associated virus vector (AAV9) for ML IV (NCT07398872) does not apply to this class [33].

7.14 Investigational therapies and developments (October 2025 to October 2026)

None of these strategies has reached human trials. The clinical developments of the same period are descriptive: recognition of cardiac-predominant forms, diagnosis from the placenta, pneumothorax as a late complication of ML II and cohorts from Turkey, China, Egypt and Iran [58, 59, 63, 72, 77, 78, 87, 93, 95, 99].

8. Reproductive Options and Family Counselling

8.1 Recurrence risks

8.2 Carrier detection and cascade testing

Carrier detection requires prior identification of the familial variants, because enzyme levels do not distinguish carriers from non-carriers [7, 44]. In a population isolate with several at-risk couples, Vidgoff and colleagues (1982) proposed identifying ML II carriers by serum beta-hexosaminidase B; this historical approach was not adopted as a carrier test outside that context [15]. The Atlas notes that heterozygotes have, on average, intermediate phosphotransferase activity in leucocytes and fibroblasts, but this difference does not allow reliable individual classification [40, 44]. Cascade testing extends to adult siblings, uncles and aunts and cousins, and testing of the partner is indicated, especially when consanguinity is likely or the ethnic origin is the same [1, 7]. In populations with founder variants (Saguenay–Lac-Saint-Jean, Irish Travellers, Druze and Bedouin of northern Israel, Sephardi Jews from Tunisia), the carrier frequency justifies targeted testing of the partner for the founder variant and is an argument for population carrier screening [1, 7, 23]. In Brazil, the frequency of c.3503_3504del (11 of 24 alleles in a series of 12 patients) led to the proposal of exon 19 as the starting point for analysis [71].

8.3 Prenatal diagnosis and preimplantation genetic testing

Once the familial variants have been identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible; molecular diagnosis is preferable to biochemical diagnosis [1, 44, 73]. Without known variants, assay of several lysosomal enzymes in amniotic fluid or of phosphotransferase in villi allows diagnosis in the first trimester [2, 40]. Ultrasound may show skeletal dysplasia, fractures or hydrops, but is not abnormal in all affected fetuses, and an ultrasound phenotype of skeletal dysplasia in a pregnancy without a previous diagnosis led to identification of a novel GNPTAB variant useful for counselling [44, 88]. Examination of the placenta may provide the diagnosis after delivery, and a dichorionic twin pregnancy with only one affected fetus has been described [99, 100]. Professionals and families may have different perspectives on prenatal diagnosis, especially in the attenuated forms, and the discussion should be non-directive [1].

8.4 Content of counselling

Counselling explains the nature of the disease, inheritance, variability between ML II, intermediate and ML III, the meaning of a VUS and the importance of confirming the phase of the variants in the parents [1, 7]. The optimal time to assess genetic risk and discuss prenatal diagnosis is before pregnancy, and GeneReviews recommends offering counselling to young adults who are affected, carriers or at risk of being carriers [1, 7]. Inborn Metabolic Diseases recommends that all families be referred for genetic counselling and carrier detection and informed about prenatal and preimplantation diagnosis [31]. Identification of GNPTAB variants in an adult with cardiomyopathy implies counselling of siblings and offspring, even without a skeletal phenotype [58, 95].

9. Special Considerations: Population Diversity and Epidemiology

9.1 Frequency and what each estimate measures

The frequency estimates for ML II/III are of different kinds: birth prevalence calculated from laboratory diagnoses and the number of live births in a period, incidences in founder populations, derived estimates and records in the Orphanet database. They are not comparable with each other and do not represent current national prevalences.

The Orphanet records for ML III contain inconsistent values: the incidence attributed to Portugal (1.68/100,000) does not match that of the original article (1.89/100,000), and the point prevalences given for Europe, the Netherlands and Portugal (class 1–5/10,000) and for ML III alpha/beta worldwide contradict the GeneReviews chapter cited as the source; these values were not used [1, 18, 19, 21]. The absence of estimates for the USA and Brazil in the sources consulted does not mean that such data do not exist.

9.2 Europe and founder populations

Portugal has the highest birth prevalence of ML II/III among European studies: the study by Pinto and colleagues, based on the enzymology laboratory of the Instituto de Genética Médica Jacinto de Magalhães (Porto), the only one in Portugal performing postnatal and prenatal diagnosis of lysosomal diseases at the time, calculated an overall prevalence of lysosomal diseases of 25/100,000 live births in the north of the country, about twice that of the Netherlands and Australia, and the mucolipidoses were among the diseases with a higher prevalence than in those populations [21]. Portuguese research has contributed to the molecular characterisation of the class: Encarnação’s series of 13 patients (10 Portuguese), the functional study of the three Afghan siblings with the GNPTG intronic variant IVS7-10G>A, the study of the origin of c.3503_3504del (with 5 Portuguese patients), the 2019 worldwide variant review and the development of antisense oligonucleotides [53, 70, 76, 80, 137, 138]. In Encarnação’s series, c.3503_3504del accounted for 45% of the mutant alleles of patients with ML II, and the Portuguese patient with ML III gamma in that series had onset at 10 years and reached adulthood [11, 13].

In Ireland, ML II is concentrated in the Irish Traveller community, and in the Irish series 3 of the 5 patients with hyperparathyroidism were children of Irish Traveller parents [1, 22]. In Saguenay–Lac-Saint-Jean, the carrier frequency of 1/39 is the highest documented [23]. The c.3503_3504del variant has a common haplotype compatible with a peri-Mediterranean origin about 2000 years ago and is present in Italy, Portugal, Turkey, Israel and Palestine, Canada, Argentina, Brazil and Ireland [70]. Most of the patients with ML III gamma known to the GeneReviews authors are concentrated in the Mediterranean region, with the c.499dup variant in the Druze and Bedouin of Israel, Sephardi Jews from Tunisia and in Turkey [7]. In Turkey, consanguinity was present in 63.1% of the 19 children at a tertiary centre, and c.3503_3504del was the most common variant in ML II [63].

9.3 The Americas

In the USA, the main clinical series is that of the Greenwood Genetic Center (61 probands) [24]. In Brazil, a retrospective multicentre study (1983–2013) brought together 27 patients (15 ML II, 9 ML III alpha/beta, 3 ML III gamma), with a median age at diagnosis of 2.7 years and a predominantly skeletal presentation [86]; a Brazilian series described the radiological findings of 5 patients referred for disproportionate short-trunk short stature and waddling gait [108]; molecular analysis of 12 patients identified c.3503_3504del in 11 of 24 alleles [71]; and a Brazilian family with ML III gamma documented an allele arising de novo or through maternal germline mosaicism [107]. GeneReviews includes South America among the regions of origin of patients with ML III gamma identified by next-generation sequencing [7].

9.4 Asia, the Middle East and North Africa

In Japan, 40 patients (25 ML II and 15 ML III) showed recurrent variants distinct from the European ones [67]. In China, the review of 56 patients with ML II identified c.1090C>T p.(Arg364Ter) in 39.3%, associated with early death [72]. In Korea no hotspot was identified [73]. In India and Turkey the largest cohort of ML III gamma was described, all homozygous, with consanguinity in 8 of 12 families [28]. Recent series from Egypt, Yemen and Iran illustrate the contribution of consanguinity and private variants [77, 78]. In Morocco, ML III gamma was described as a familial scleroderma-like form [96].

9.5 Special populations

In populations with high consanguinity, homozygosity for private variants is the rule and partner testing should be offered to all at-risk couples [7, 28]. Children with ML II and people with ML III undergoing anaesthesia constitute a population with a specific perioperative risk (7.8). Adults with undiagnosed ML III may be followed in cardiology, rheumatology or orthopaedics, and genome sequencing has brought cardiac-predominant forms to light [27, 58].

9.6 European network and patient organisations

The relevant European reference network is MetabERN (hereditary metabolic disorders), which includes lysosomal diseases; ERN BOND (rare bone diseases) covers the skeletal component [36, 37]. The patient organisations indicated by the GeneReviews chapters and by the data collection for this article are Cure Mucolipidosis (curemucolipidosis.org), the International Society for Mannosidosis & Related Diseases (ISMRD, ismrd.org), which presents itself as the international advocate for glycoprotein storage diseases and includes ML II and ML III gamma, Vaincre les Maladies Lysosomales (France, vml-asso.org), the MPS Society in the United Kingdom (mpssociety.org.uk) and the National MPS Society in the USA (mpssociety.org) [1, 7, 149]. OrphanAnesthesia publishes the anaesthesia recommendation in open access [92].

10. Illustrative Clinical Case

10.1 Hand stiffness from 6 months: early-onset ML III gamma (La Rosa et al., 2025)

A girl, the second child of non-consanguineous Caucasian parents from a small village on an Italian island, had a normal pregnancy and developmental milestones [14]. At 6 months, the parents noticed abnormal flexion of the second fingers when the child crawled, which progressed to inability to make a fist and grasp objects [14]. At 4 years, radiographs and ultrasound of the hands were reported as normal; on later review, the radiographs showed broad, short metacarpals, brachymetacarpia of the third and fourth digits bilaterally and distal widening of the radius, mild signs of dysostosis multiplex (figure 1 of the article, hand radiographs at 4 years alongside a control: https://pmc.ncbi.nlm.nih.gov/articles/PMC12363594/figure/f0005/) [14]. At 4.5 years, at a paediatric rheumatology centre, genetic testing for Farber disease, the Berry test for urinary GAGs and ophthalmological and cardiological assessments were negative or inconclusive [14].

At 5.7 years, the child weighed 21.8 kg (75th centile) and measured 114 cm (50th centile), with a parental target height of 163.5 ± 8.5 cm; growth had followed her centiles [14]. Examination showed claw hand, swelling of the fingers, a short neck, class III malocclusion and minimal limitation of the temporomandibular joint, without visceromegaly or neurological changes [14]. The radiographs (figure 3 of the article: https://pmc.ncbi.nlm.nih.gov/articles/PMC12363594/figure/f0015/) showed short, broad metacarpals, “bullet-shaped” deformity of the proximal interphalangeal joints, anterior beaking of the lumbar vertebral bodies, mild platyspondyly and hyperlordosis [14]. Echocardiography showed mild thickening of the mitral leaflets with normal flow, and ophthalmological examination showed no corneal clouding, only hypermetropic anisometropia [14]. JIA and progressive pseudorheumatoid arthropathy were excluded [14].

Leucocyte alpha-L-iduronidase and arylsulfatase B and urinary GAGs by the DMB method were normal, which excluded MPS I and MPS VI; in plasma, total beta-hexosaminidase was 7868 nmol/ml/h (reference 883 ± 252) and alpha-L-fucosidase 1177 nmol/ml/h (reference 595 ± 193) [14]. Study of GNPTG identified compound heterozygosity for NM_032520.5:c.110+1G>A, of paternal origin, and c.654_657del p.(Tyr218Ter) (in the authors’ notation, c.654_657delCTTA), of maternal origin, and the diagnosis was early-onset ML III gamma [10, 14]. The family received genetic counselling and no other relatives with suggestive symptoms were identified [14]. At 8 years, the child continued with regular visits, an intensive rehabilitation programme and orthoses to improve mobility of the interphalangeal and carpal joints, with no aetiological treatment available [14].

The case shows the unusually early onset for a form whose median onset is 3.0 years, the delay of about 5 years between the first sign and diagnosis, the passage through rheumatology and the value of the plasma hydrolase panel when urinary GAGs and leucocyte enzymes are normal [7, 14, 87]. The term “intensive rehabilitation” should be read in the light of the GeneReviews recommendation of low-impact therapies and avoidance of vigorous stretching, which may damage capsules and tendons [7]. The authors propose a flowchart for the investigation of childhood joint stiffness that includes lysosomal diseases (figure 4 of the article, not consulted as an image) [14].

10.2 Arrhythmogenic cardiomyopathy as the presentation of ML III alpha/beta (Castrichini et al., 2026)

A 38-year-old man with no previous history attended the emergency department with chest pain and palpitations [58]. Troponin was normal; the electrocardiogram showed low voltage with extensive T-wave inversion and frequent ventricular extrasystoles; coronary angiography was normal; echocardiography showed systolic dysfunction with apical hypokinesia; cardiac magnetic resonance showed a left ventricle of normal dimensions and thickness, apical dyskinesia, an ejection fraction of 34% and almost circumferential “ring-like” subepicardial late enhancement, without oedema, and the findings met the Padua criteria for arrhythmogenic left ventricular cardiomyopathy [58]. Because of the extensive fibrosis and the arrhythmic burden, a cardioverter-defibrillator was implanted for primary prevention [58]. A commercial 168-gene cardiac panel was negative; genome sequencing of the patient and his parents identified, in GNPTAB, c.771G>A p.(Leu257=), a splice variant of maternal origin in exon 7, previously described in an adult-onset form of ML III with cardiomyopathy and sensory neuropathy, and c.2502del p.(Lys834AsnfsTer5), of paternal origin, confirmed by Sanger sequencing [10, 57, 58]. Metabolic assessment showed normal urinary and blood lysosomal biomarkers and enzyme activity, an unremarkable skeleton, normal cognition, subtly coarse facial features, mild corneal opacity and cataracts [58]. Shortly after implantation, the patient had an episode of ventricular tachycardia terminated by antitachycardia pacing and, later, a ventricular tachycardia storm with syncope and multiple shocks, which prompted the start of amiodarone [58].

The case shows that normal hydrolases and a normal skeleton do not exclude the disease when the genotype includes a splice allele with residual activity, and that cardiomyopathy panels without GNPTAB do not detect it [58]. The arrhythmia management described (cardioverter-defibrillator and amiodarone) is that of arrhythmogenic cardiomyopathy and not a recommendation specific to ML.

11. Prognosis and Long-Term Follow-Up

11.1 Prognosis

In the review of 843 published cases, median survival was 5.0 years in ML II and 62.0 years in ML III, and median age at death 1.8 and 33.0 years, respectively; pulmonary and cardiac complications were the most frequent causes of death in all forms [25]. These values derive from published cases, with publication bias, and patients described after 2000 had significantly better survival than those described before 1980 [7, 25]. In recent centre series, median survival in ML II was 28 months in Turkey and mean survival 36 months in Manchester, and 4 of 5 Irish children with neonatal hyperparathyroidism died between 10 months and 7 years [22, 63, 68]. In China, the c.1090C>T variant was associated with early death [72]. The intermediate form has a life expectancy similar to that of ML III, with death mainly from valvular and myocardial disease [1]. In ML III alpha/beta, life expectancy data are scarce, the oldest known patient was 86 years old and death in early or mid-adulthood is often cardiopulmonary [1, 5]. In ML III gamma, there are no specific life expectancy data and a slight reduction is presumed [7]. The morbidity of ML III is dominated by chronic pain, loss of mobility and repeated surgery: 46% of Oussoren’s adults depended on help or a wheelchair [27].

11.2 Surveillance table

The GNPTAB GeneReviews does not set intervals for echocardiography, densitometry and ophthalmology in ML III alpha/beta; the annual and 2- to 5-year intervals of the ML III gamma GeneReviews are the only ones published for ML III, and their application to the alpha/beta form is an extrapolation, all the more so as bone and cardiac disease is generally more severe in the alpha/beta form [1, 7, 56]. The ML III gamma GeneReviews summarises surveillance as an annual assessment, unless pain or cardiac or respiratory manifestations require more [7].

11.3 Assessment of at-risk relatives

Siblings of a proband with ML III should be assessed clinically and molecularly, because the attenuated form may go unnoticed for years, and identification of an adult with cardiac-predominant ML III implies cardiac and molecular assessment of siblings [1, 58, 95]. Carriers do not need surveillance [1].

12. Summary and Concluding Remarks

Mucolipidoses II and III are autosomal recessive lysosomal diseases caused by deficiency of GlcNAc-1-phosphotransferase, the enzyme that adds the mannose-6-phosphate marker to lysosomal hydrolases. The alpha and beta subunits are encoded by GNPTAB and the gamma subunit by GNPTG. Without the marker, the hydrolases are secreted: their activity rises in plasma and is lacking inside mesenchymal cells, which explains the coexistence of dysostosis multiplex, joint disease, valvular disease and thickening of connective tissues and airways.

The phenotype follows residual activity. ML II alpha/beta, with complete loss of activity, is evident at birth, with gingival hyperplasia, thick skin, contractures, bone changes of transient hyperparathyroidism and progressive cardiorespiratory disease; median survival in published cases is 5 years. The intermediate form associated with c.10A>C p.(Lys4Gln) combines the physical appearance of ML II with the development and survival of ML III. ML III alpha/beta manifests at around 3 years with stiffness of the shoulders, hips and fingers, carpal tunnel, destructive hip disease, painful osteoporosis and valvular disease, and is often confused with attenuated MPS or juvenile idiopathic arthritis. Cardiac-predominant forms exist in adults, sometimes with a normal skeleton. ML III gamma is generally milder, with normal bone remodelling, and its distinction from the alpha/beta form requires molecular testing.

Diagnosis rests on elevation of several hydrolases in plasma and identification of biallelic variants in GNPTAB or GNPTG; urinary tests may be normal and sequencing should be complemented by deletion analysis, parental testing and, when the phenotype is atypical, genome sequencing and RNA studies. There is no disease-modifying therapy. Haematopoietic stem cell transplantation has not improved survival or skeletal progression in ML II, and bisphosphonates, used off-label with doses extrapolated from osteogenesis imperfecta, relieve pain in some patients with GNPTAB-related ML III. Treatment is supportive, with low-impact therapies, orthopaedic surgery at the appropriate time, cardiac and respiratory surveillance, anaesthetic planning in a tertiary centre because of the difficult airway and endocarditis prophylaxis according to the cardiology guidelines. Experimental strategies (gene therapy, antisense oligonucleotides, delivery of mannose-6-phosphate-bearing proteins, zoledronic acid in animal models) remain preclinical, and the exon 19 skipping strategy has proved ineffective. Portugal has the highest birth prevalence of ML II/III among European studies, and the founder populations of Quebec, the Irish Travellers and the Druze and Bedouin of Israel justify targeted carrier testing.

13. References

1. Erdem F, Canda E. GNPTAB-Related Disorders. 2008 Aug 26 [Updated 2026 Aug 13]. 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/NBK1828/
2. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Mucolipidosis II (MIM 252500). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 124–128.
3. Saul RA, Proud V, Taylor HA, et al. Prenatal mucolipidosis type II (I-cell disease) can present as Pacman dysplasia. Am J Med Genet A. 2005;135:328–32. doi:10.1002/ajmg.a.30716. PMID: 15887289.
4. Leroy JG, Sillence D, Wood T, Barnes J, Lebel RR, Friez MJ, et al. A novel intermediate mucolipidosis II/IIIαβ caused by GNPTAB mutation in the cytosolic N-terminal domain. Eur J Hum Genet. 2014;22(5):594-601. doi:10.1038/ejhg.2013.207. PMID: 24045841. PMC3992569.
5. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Mucolipidosis III (MIM 252600, 252605). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 129–134.
6. Cathey SS, Kudo M, Tiede S, Raas-Rothschild A, Braulke T, Beck M, et al. Molecular order in mucolipidosis II and III nomenclature. Am J Med Genet A. 2008;146A(4):512-3. doi:10.1002/ajmg.a.32193. PMID: 18203164.
7. Raas-Rothschild A, Spiegel R. Mucolipidosis III Gamma. 2010 Jan 28 [Updated 2026 Apr 23]. 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/NBK24701/
8. OMIM® Online Mendelian Inheritance in Man. #252605 MUCOLIPIDOSIS III GAMMA. Johns Hopkins University; last edited 21/01/2025. https://omim.org/entry/252605. Accessed 28 October 2026.
9. GeneGuidance. GNPTAB — clinical curation of the gene and associated conditions (ML II alpha/beta, OMIM 252500; ML III alpha/beta, OMIM 252600). Version of 26 September 2026.
10. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. PMID: 28967166. PMC5765404. Tool: https://variantvalidator.org (GRCh38, MANE Select; accessed 27 and 28 October 2026).
11. OMIM® Online Mendelian Inheritance in Man. *607840 N-ACETYLGLUCOSAMINE-1-PHOSPHOTRANSFERASE, ALPHA/BETA SUBUNITS; GNPTAB. Johns Hopkins University; last edited 21/08/2026. https://omim.org/entry/607840. Accessed 28 October 2026.
12. GeneGuidance. GNPTG — clinical curation of the gene and associated condition (ML III gamma, OMIM 252605). Version of 26 September 2026.
13. OMIM® Online Mendelian Inheritance in Man. *607838 N-ACETYLGLUCOSAMINE-1-PHOSPHOTRANSFERASE, GAMMA SUBUNIT; GNPTG. Johns Hopkins University; last edited 31/07/2023. https://omim.org/entry/607838. Accessed 28 October 2026.
14. La Rosa A, Pepe A, Tappino B, Corsolini F, Chiaro A, Madeo A. Hand stiffness not only a rheumatological sign: A case of early onset mucolipidosis III-gamma with literature review. Mol Genet Metab Rep. 2025;44:101246. doi:10.1016/j.ymgmr.2025.101246. PMID: 40838094. PMC12363594.
15. OMIM® Online Mendelian Inheritance in Man. #252500 MUCOLIPIDOSIS II ALPHA/BETA. Johns Hopkins University; last edited 21/08/2026. https://omim.org/entry/252500. Accessed 27 October 2026.
16. OMIM® Online Mendelian Inheritance in Man. #252600 MUCOLIPIDOSIS III ALPHA/BETA. Johns Hopkins University; last edited 01/10/2024. https://omim.org/entry/252600. Accessed 28 October 2026.
17. Orphanet. Mucolipidosis type II (ORPHA:576). https://www.orpha.net/en/disease/detail/576
18. Orphanet. Mucolipidosis type III (ORPHA:577). https://www.orpha.net/en/disease/detail/577
19. Orphanet. Mucolipidosis type III alpha/beta (ORPHA:423461). https://www.orpha.net/en/disease/detail/423461
20. Orphanet. Mucolipidosis type III gamma (ORPHA:423470). https://www.orpha.net/en/disease/detail/423470
21. Pinto R, Caseiro C, Lemos M, Lopes L, Fontes A, Ribeiro H, et al. Prevalence of lysosomal storage diseases in Portugal. Eur J Hum Genet. 2004;12(2):87-92. doi:10.1038/sj.ejhg.5201044. PMID: 14685153.
22. Boruah R, Monavari AA, Conlon T, Murphy N, Stroiescu A, Ryan S, et al. Secondary Hyperparathyroidism in Children with Mucolipidosis Type II (I-Cell Disease): Irish Experience. J Clin Med. 2022;11(5). doi:10.3390/jcm11051366. PMID: 35268460. PMC8911139.
23. Plante M, Claveau S, Lepage P, Lavoie EM, Brunet S, Roquis D, et al. Mucolipidosis II: a single causal mutation in the N-acetylglucosamine-1-phosphotransferase gene (GNPTAB) in a French Canadian founder population. Clin Genet. 2008;73(3):236-44. doi:10.1111/j.1399-0004.2007.00954.x. PMID: 18190596.
24. Cathey SS, Leroy JG, Wood T, Eaves K, Simensen RJ, Kudo M, et al. Phenotype and genotype in mucolipidoses II and III alpha/beta: a study of 61 probands. J Med Genet. 2010;47(1):38-48. doi:10.1136/jmg.2009.067736. PMID: 19617216. PMC3712854.
25. Dogterom EJ, Wagenmakers MAEM, Wilke M, Demirdas S, Muschol NM, Pohl S, et al. Mucolipidosis type II and type III: a systematic review of 843 published cases. Genet Med. 2021;23(11):2047-2056. doi:10.1038/s41436-021-01244-4. PMID: 34172897.
26. Tüysüz B, Kasapçopur Ö, Alkaya DU, Şahin S, Sözeri B, Yeşil G. Mucolipidosis type III gamma: Three novel mutation and genotype-phenotype study in eleven patients. Gene. 2018;642:398-407. doi:10.1016/j.gene.2017.11.052. PMID: 29170090.
27. Oussoren E, van Eerd D, Murphy E, Lachmann R, van der Meijden JC, Hoefsloot LH, et al. Mucolipidosis type III, a series of adult patients. J Inherit Metab Dis. 2018;41(5):839-848. doi:10.1007/s10545-018-0186-z. PMID: 29704188. PMC6133174.
28. Nampoothiri S, Elcioglu NH, Koca SS, Yesodharan D, Kk C, Krishnan V 5th, et al. Does the clinical phenotype of mucolipidosis-IIIγ differ from its αβ counterpart?: supporting facts in a cohort of 18 patients. Clin Dysmorphol. 2019;28(1):7-16. doi:10.1097/MCD.0000000000000249. PMID: 30507725.
29. Hong X, Pollard L, He M, Gelb MH, Wood TC. Multiplex tandem mass spectrometry enzymatic activity assay for the screening and diagnosis of mucolipidosis type II and III. Mol Genet Metab Rep. 2023;35:100978. doi:10.1016/j.ymgmr.2023.100978. PMID: 37275682.
30. Lund TC, Cathey SS, Miller WP, Eapen M, Andreansky M, Dvorak CC, et al. Outcomes after hematopoietic stem cell transplantation for children with I-cell disease. Biol Blood Marrow Transplant. 2014;20(11):1847-51. doi:10.1016/j.bbmt.2014.06.019. PMID: 25016194. PMC4194244.
31. Jones S, Wijburg FA. Glycosaminoglycans and oligosaccharides disorders: glycosaminoglycans synthesis defects, mucopolysaccharidoses, oligosaccharidoses and mucolipidoses. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 41.
32. Robinson C, Baker N, Noble J, King A, David G, Sillence D, et al. The osteodystrophy of mucolipidosis type III and the effects of intravenous pamidronate treatment. J Inherit Metab Dis. 2002;25(8):681-93. doi:10.1023/a:1022935115323. PMID: 12705498.
33. ClinicalTrials.gov. Pesquisa «mucolipidosis OR GNPTAB OR GNPTG OR I-cell disease». Available from: https://clinicaltrials.gov (accessed 2026 Oct 27).
34. European Medicines Agency. Medicines output: orphan designations, updated 7 October 2026. Available from: https://www.ema.europa.eu/en/documents/report/medicines-output-orphan_designations-json-report_en.json (accessed 2026 Oct 27).
35. U.S. Food and Drug Administration, Office of Orphan Products Development. Search Orphan Drug Designations and Approvals: searches by designation “mucolipidosis” and “I-cell”. Available from: https://www.accessdata.fda.gov/scripts/opdlisting/oopd/ (accessed 2026 Oct 27 and 2026 Oct 28).
36. MetabERN — European Reference Network for Hereditary Metabolic Disorders. Available from: https://metab.ern-net.eu/
37. ERN BOND — European Reference Network on Rare Bone Diseases. Available from: https://ernbond.eu/
38. Kang C, Riazuddin S, Mundorff J, Krasnewich D, Friedman P, Mullikin JC, et al. Mutations in the lysosomal enzyme-targeting pathway and persistent stuttering. N Engl J Med. 2010;362(8):677-85. doi:10.1056/NEJMoa0902630. PMID: 20147709. PMC2936507.
39. Raza MH, Domingues CE, Webster R, Sainz E, Paris E, Rahn R, et al. Mucolipidosis types II and III and non-syndromic stuttering are associated with different variants in the same genes. Eur J Hum Genet. 2016;24(4):529-34. doi:10.1038/ejhg.2015.154. PMID: 26130485. PMC4929873.
40. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Mucolipidosis II and III (I-cell disease and pseudo-Hurler polydystrophy) N-acetyl-glucosaminyl-1-phosphotransferase deficiency. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 615–622.
41. Misko A, Grishchuk Y, Goldin E, Schiffmann R. Mucolipidosis IV. 2005 Jan 28 [Updated 2021 Feb 11]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1214/
42. Maroteaux P, Lamy M. La pseudo-polydystrophie de Hurler. Presse Med. 1966;74:2889–92. PMID: 4958870.
43. Leroy JG, Demars RI. Mutant enzymatic and cytological phenotypes in cultured human fibroblasts. Science. 1967;157(3790):804-6. doi:10.1126/science.157.3790.804. PMID: 17842782.
44. Jones KL, Jones MC, del Campo M. Mucolipidosis II (Leroy I-cell disease). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 638–641.
45. Leroy JG, Spranger JW, Feingold M, Opitz JM, Crocker AC. I-cell disease: a clinical picture. J Pediatr. 1971;79(3):360-5. doi:10.1016/s0022-3476(71)80142-7. PMID: 4327936.
46. Hickman S, Neufeld EF. A hypothesis for I-cell disease: defective hydrolases that do not enter lysosomes. Biochem Biophys Res Commun. 1972;49(4):992-9. doi:10.1016/0006-291x(72)90310-5. PMID: 4345092.
47. Reitman ML, Varki A, Kornfeld S. Fibroblasts from patients with I-cell disease and pseudo-Hurler polydystrophy are deficient in uridine 5'-diphosphate-N-acetylglucosamine: glycoprotein N-acetylglucosaminylphosphotransferase activity. J Clin Invest. 1981;67(5):1574-9. doi:10.1172/jci110189. PMID: 6262380. PMC370727.
48. Reitman ML, Kornfeld S. UDP-N-acetylglucosamine:glycoprotein N-acetylglucosamine-1-phosphotransferase. Proposed enzyme for the phosphorylation of the high mannose oligosaccharide units of lysosomal enzymes. J Biol Chem. 1981;256(9):4275-81. PMID: 6452459.
49. Varki AP, Reitman ML, Kornfeld S. Identification of a variant of mucolipidosis III (pseudo-Hurler polydystrophy): a catalytically active N-acetylglucosaminylphosphotransferase that fails to phosphorylate lysosomal enzymes. Proc Natl Acad Sci U S A. 1981;78(12):7773-7. doi:10.1073/pnas.78.12.7773. PMID: 6461005. PMC349353.
50. Raas-Rothschild A, Cormier-Daire V, Bao M, Genin E, Salomon R, Brewer K, et al. Molecular basis of variant pseudo-hurler polydystrophy (mucolipidosis IIIC). J Clin Invest. 2000;105(5):673-81. doi:10.1172/JCI5826. PMID: 10712439. PMC289169.
51. Tiede S, Storch S, Lübke T, Henrissat B, Bargal R, Raas-Rothschild A, et al. Mucolipidosis II is caused by mutations in GNPTA encoding the alpha/beta GlcNAc-1-phosphotransferase. Nat Med. 2005;11(10):1109-12. doi:10.1038/nm1305. PMID: 16200072.
52. Kudo M, Brem MS, Canfield WM. Mucolipidosis II (I-cell disease) and mucolipidosis IIIA (classical pseudo-hurler polydystrophy) are caused by mutations in the GlcNAc-phosphotransferase alpha / beta -subunits precursor gene. Am J Hum Genet. 2006;78(3):451-63. doi:10.1086/500849. PMID: 16465621. PMC1380288.
53. Velho RV, Harms FL, Danyukova T, Ludwig NF, Friez MJ, Cathey SS, et al. The lysosomal storage disorders mucolipidosis type II, type III alpha/beta, and type III gamma: Update on GNPTAB and GNPTG mutations. Hum Mutat. 2019;40(7):842-864. doi:10.1002/humu.23748. PMID: 30882951.
54. Disorders of cellular trafficking. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 44.
55. Bargal R, Zeigler M, Abu-Libdeh B, Zuri V, Mandel H, Ben Neriah Z, et al. When Mucolipidosis III meets Mucolipidosis II: GNPTA gene mutations in 24 patients. Mol Genet Metab. 2006;88(4):359-63. doi:10.1016/j.ymgme.2006.03.003. PMID: 16630736.
56. Di Lorenzo G, Westermann LM, Yorgan TA, Stürznickel J, Ludwig NF, Ammer LS, et al. Pathogenic variants in GNPTAB and GNPTG encoding distinct subunits of GlcNAc-1-phosphotransferase differentially impact bone resorption in patients with mucolipidosis type II and III. Genet Med. 2021;23(12):2369-2377. doi:10.1038/s41436-021-01285-9. PMID: 34341521. PMC8629757.
57. Steet RA, Hullin R, Kudo M, Martinelli M, Bosshard NU, Schaffner T, et al. A splicing mutation in the alpha/beta GlcNAc-1-phosphotransferase gene results in an adult onset form of mucolipidosis III associated with sensory neuropathy and cardiomyopathy. Am J Med Genet A. 2005;132A(4):369-75. doi:10.1002/ajmg.a.30498. PMID: 15633164.
58. Castrichini M, Kim ET, Tester DJ, Schimmenti LA, Ackerman MJ, Giudicessi JR. Genome sequencing identifies an attenuated, cardiac-predominant form of mucolipidosis type IIIα/β as the potential cause of a genetically elusive, ring-like arrhythmogenic non-dilated left ventricular cardiomyopathy. HeartRhythm Case Rep. 2026;12(9):1128-1130. doi:10.1016/j.hrcr.2026.06.016. PMID: 42774235. PMC13593553.
59. Wu X, Song L, Kang L. Isolated cardiac mucolipidosis type III α/β mimicking Fabry's disease: a case report highlighting the role of pathology and genetics in diagnosis. Eur Heart J. 2026;47(24):3224. doi:10.1093/eurheartj/ehag233. PMID: 42048139.
60. Khan SA, Tomatsu SC. Mucolipidoses Overview: Past, Present, and Future. Int J Mol Sci. 2020;21(18). doi:10.3390/ijms21186812. PMID: 32957425. PMC7555117.
61. Li H, Doray B, Jennings BC, Lee WS, Liu L, Kornfeld S, et al. Structure of a truncated human GlcNAc-1-phosphotransferase variant reveals the basis for its hyperactivity. J Biol Chem. 2024;300(9):107706. doi:10.1016/j.jbc.2024.107706. PMID: 39178950. PMC11418123.
62. van Meel E, Qian Y, Kornfeld SA. Mislocalization of phosphotransferase as a cause of mucolipidosis III αβ. Proc Natl Acad Sci U S A. 2014;111:3532–7. doi:10.1073/pnas.1401417111. PMID: 24550498.
63. Erdem F, Canda E, Yazıcı H, Eser R, Yoldaş Çelik M, Keşan S, et al. Mucolipidosis type II and III: clinical spectrum, genetic landscape, and longitudinal outcomes in a pediatric cohort with six novel mutations. J Pediatr Endocrinol Metab. 2025;38(12):1286-1298. doi:10.1515/jpem-2025-0352. PMID: 41064848.
64. Kerr DA, Memoli VA, Cathey SS, Harris BT. Mucolipidosis type III α/β: the first characterization of this rare disease by autopsy. Arch Pathol Lab Med. 2011;135(4):503-10. doi:10.5858/2010-0236-CR.1. PMID: 21466370. PMC4188553.
65. Kollmann K, Pestka JM, Kühn SC, et al. Decreased bone formation and increased osteoclastogenesis cause bone loss in mucolipidosis II. EMBO Mol Med. 2013;5:1871–86. doi:10.1002/emmm.201302979. PMID: 24127423.
66. Unger S, Paul DA, Nino MC, et al. Mucolipidosis II presenting as severe neonatal hyperparathyroidism. Eur J Pediatr. 2005;164:236–43. doi:10.1007/s00431-004-1591-x. PMID: 15580357.
67. Otomo T, Muramatsu T, Yorifuji T, Okuyama T, Nakabayashi H, Fukao T, et al. Mucolipidosis II and III alpha/beta: mutation analysis of 40 Japanese patients showed genotype-phenotype correlation. J Hum Genet. 2009;54(3):145-51. doi:10.1038/jhg.2009.3. PMID: 19197337.
68. Edmiston R, Wilkinson S, Jones S, Tylee K, Broomfield A, Bruce IA. I-Cell Disease (Mucolipidosis II): A Case Series from a Tertiary Paediatric Centre Reviewing the Airway and Respiratory Consequences of the Disease. JIMD Rep. 2019;45:1-8. doi:10.1007/8904_2018_130. PMID: 30209781. PMC6336676.
69. National Center for Biotechnology Information. ClinVar: VCV000002771, VCV000002774, VCV000684627, VCV000039021, VCV000038413, VCV000068107, VCV000437454. Available from: https://www.ncbi.nlm.nih.gov/clinvar/ (accessed 2026 Oct 27).
70. Coutinho MF, Encarnação M, Gomes R, da Silva Santos L, Martins S, Sirois-Gagnon D, et al. Origin and spread of a common deletion causing mucolipidosis type II: insights from patterns of haplotypic diversity. Clin Genet. 2011;80(3):273-80. doi:10.1111/j.1399-0004.2010.01539.x. PMID: 20880125.
71. Cury GK, Matte U, Artigalás O, Alegra T, Velho RV, Sperb F, et al. Mucolipidosis II and III alpha/beta in Brazil: analysis of the GNPTAB gene. Gene. 2013;524(1):59-64. doi:10.1016/j.gene.2013.03.105. PMID: 23566849.
72. Gan X, Li J, Cao J. Clinical phenotype and GNPTAB gene mutation spectrum analysis of Chinese patients with mucolipidosis type II α/β. Front Med (Lausanne). 2026;13:1895262. doi:10.3389/fmed.2026.1895262. PMID: 42553122. PMC13433197.
73. Yang M, Cho SY, Park HD, Choi R, Kim YE, Kim J, et al. Clinical, biochemical and molecular characterization of Korean patients with mucolipidosis II/III and successful prenatal diagnosis. Orphanet J Rare Dis. 2017;12(1):11. doi:10.1186/s13023-016-0556-2. PMID: 28095893. PMC5240260.
74. Paik KH, Song SM, Ki CS, Yu HW, Kim JS, Min KH, et al. Identification of mutations in the GNPTA (MGC4170) gene coding for GlcNAc-phosphotransferase alpha/beta subunits in Korean patients with mucolipidosis type II or type IIIA. Hum Mutat. 2005;26(4):308-14. doi:10.1002/humu.20205. PMID: 16116615.
75. Mao SJ, Zu YM, Dai YL, Zou CC. Case Report: Mucolipidosis II and III Alpha/Beta Caused by Pathogenic Variants in the GNPTAB Gene (Mucolipidosis). Front Pediatr. 2022;10:852701. doi:10.3389/fped.2022.852701. PMID: 35463894. PMC9033252.
76. Encarnação M, Lacerda L, Costa R, Prata MJ, Coutinho MF, Ribeiro H, et al. Molecular analysis of the GNPTAB and GNPTG genes in 13 patients with mucolipidosis type II or type III - identification of eight novel mutations. Clin Genet. 2009;76(1):76-84. doi:10.1111/j.1399-0004.2009.01185.x. PMID: 19659762.
77. Eissa NR, Abdel-Salam GM, Kretkiewicz M, Pieters E, Jaeken J, Matthijs G, et al. Mucolipidosis II: novel variants, clinical variation and assessment of HAP1 cells as a disease model. Hum Mol Genet. 2026;35(17). doi:10.1093/hmg/ddag068. PMID: 42574251.
78. Fathi M, Khalilian S, Karimzadeh P, Anvari S, Alaei MR, Sadeghi B, et al. Molecular characterization of mucolipidosis: A case-series from Iran. Gene. 2026;1017:150464. doi:10.1016/j.gene.2026.150464. PMID: 42906832.
79. van Meel E, Kornfeld S. Mucolipidosis III GNPTG Missense Mutations Cause Misfolding of the γ Subunit of GlcNAc-1-Phosphotransferase. Hum Mutat. 2016;37(7):623-6. doi:10.1002/humu.22993. PMID: 27038293. PMC4907843.
80. Pohl S, Encarnacão M, Castrichini M, Müller-Loennies S, Muschol N, Braulke T. Loss of N-acetylglucosamine-1-phosphotransferase gamma subunit due to intronic mutation in GNPTG causes mucolipidosis type III gamma: Implications for molecular and cellular diagnostics. Am J Med Genet A. 2010;152A(1):124-32. doi:10.1002/ajmg.a.33170. PMID: 20034096.
81. Persichetti E, Chuzhanova NA, Dardis A, et al. Identification and molecular characterization of six novel mutations in the UDP-N-acetylglucosamine-1-phosphotransferase gamma subunit (GNPTG) gene in patients with mucolipidosis III gamma. Hum Mutat. 2009;30:978–84. doi:10.1002/humu.20959. PMID: 19370764.
82. Han TU, Root J, Reyes LD, Huchinson EB, Hoffmann JD, Lee WS, et al. Human GNPTAB stuttering mutations engineered into mice cause vocalization deficits and astrocyte pathology in the corpus callosum. Proc Natl Acad Sci U S A. 2019;116(35):17515-17524. doi:10.1073/pnas.1901480116. PMID: 31405983. PMC6717282.
83. Bishop D, Wang S, SheikhBahaei S. A stuttering-associated Gnptab variant alters fine-motor kinematics. bioRxiv. 2025. doi:10.64898/2025.12.08.693031. PMID: 41415373. PMC12710626.
84. Ammer LS, Oussoren E, Muschol NM, Pohl S, Rubio-Gozalbo ME, Santer R, et al. Hip Morphology in Mucolipidosis Type II. J Clin Med. 2020;9(3). doi:10.3390/jcm9030728. PMID: 32182687. PMC7141371.
85. Ammer LS, Täuber K, Perez A, Dohrmann T, Denecke J, Santer R, et al. CNS Manifestations in Mucolipidosis Type II-A Retrospective Analysis of Longitudinal Data on Neurocognitive Development and Neuroimaging in Eleven Patients. J Clin Med. 2023;12(12). doi:10.3390/jcm12124114. PMID: 37373807. PMC10299287.
86. Alegra T, Sperb-Ludwig F, Guarany NR, Ribeiro EM, Lourenço CM, Kim CA, et al. Clinical Characterization of Mucolipidoses II and III: A Multicenter Study. J Pediatr Genet. 2019;8(4):198-204. doi:10.1055/s-0039-1697605. PMID: 31687257. PMC6824887.
87. Burgac E, Kaplan İ, Bulut FD, Kara E, Köseci B, Tuğ Bozdoğan S, et al. Mucolipidosis Type III: A Hidden Challenge in the Differential Diagnosis of Rheumatological Disorders. Turk Arch Pediatr. 2026;61(5):399-405. doi:10.65717/TurkArchPediatr.2026.25402. PMID: 42084884.
88. Aggarwal S, Coutinho MF, Dalal AB, Mohamed Nurul Jain SJ, Prata MJ, Alves S. Prenatal skeletal dysplasia phenotype in severe MLII alpha/beta with novel GNPTAB mutation. Gene. 2014;542(2):266-8. doi:10.1016/j.gene.2014.03.053. PMID: 24685522.
89. Clinical approach to inborn errors of metabolism in paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1.
90. de Bode CJ, Dogterom EJ, Rozeboom AVJ, Langendonk JJ, Wolvius EB, van der Ploeg AT, et al. Orofacial abnormalities in mucopolysaccharidosis and mucolipidosis type II and III: A systematic review. JIMD Rep. 2022;63(6):621-629. doi:10.1002/jmd2.12331. PMID: 36341168. PMC9626671.
91. Kovacevic A, Schranz D, Meissner T, et al. Mucolipidosis II complicated by severe pulmonary hypertension. Mol Genet Metab. 2011;104:192–3. doi:10.1016/j.ymgme.2011.06.026. PMID: 21802970.
92. Roth W. Anaesthesia recommendations for patients suffering from Mucolipidosis II and III. Anästh Intensivmed. 2015;56:S636-S641 (OrphanAnesthesia; last modified Jun 2011). Available from: https://www.orphananesthesia.eu/en/rare-diseases/published-guidelines/mucolipidosis-type-2-and-3.html
93. Tomari S, Iijima H, Koinuma G, Kosuga M, Shoji K, Kubota M. Pneumothorax as a life-threatening long-term complication with mucolipidosis type II. Pediatr Int. 2026;68(1):e70318. doi:10.1111/ped.70318. PMID: 41566634.
94. Ammer LS, Muschol NM, Santer R, Lang A, Breyer SR, Sasu PB, et al. Anaesthesia-Relevant Disease Manifestations and Perianaesthetic Complications in Patients with Mucolipidosis-A Retrospective Analysis of 44 Anaesthetic Cases in 12 Patients. J Clin Med. 2022;11(13). doi:10.3390/jcm11133650. PMID: 35806935. PMC9267794.
95. Stewart N, Henden N, Smith-Díaz CC, Rigby J, Butters A, Baker A, et al. Cardiomyopathy and Sudden Cardiac Death as a Rare Presentation of Mucolipidosis Type III in a Family With Compound Heterozygous Variants in GNPTAB. Circ Genom Precis Med. 2026;19(4):e005703. doi:10.1161/CIRCGEN.126.005703. PMID: 42227110.
96. Zrhidri A, Amasdl S, Lyahyai J, Elouardi H, Chkirate B, Raymond L, et al. Next Generation Sequencing identifies mutations in GNPTG gene as a cause of familial form of scleroderma-like disease. Pediatr Rheumatol Online J. 2017;15(1):72. doi:10.1186/s12969-017-0200-2. PMID: 28950892. PMC5615433.
97. Ammer LS, Pohl S, Breyer SR, Aries C, Denecke J, Perez A, et al. Is hematopoietic stem cell transplantation a therapeutic option for mucolipidosis type II?. Mol Genet Metab Rep. 2021;26:100704. doi:10.1016/j.ymgmr.2020.100704. PMID: 33505859. PMC7815485.
98. Firth HV, Hurst JA. Oedema, nuchal translucency and cystic hygroma. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 768–769.
99. LaScala N, Rose L, Sanchez J, Chousal JN, Murray S, Parast M, et al. A Case Report of I-Cell Disease (Mucolipidosis Type II) With a Novel Placental-Focused Diagnostic Approach. Pediatr Dev Pathol. 2026:10935266261440594. doi:10.1177/10935266261440594. PMID: 42007676.
100. Chapel DB, Choy B, Pytel P, Husain AN, Lastra RR. Mucolipidosis Type II Affecting 1 Fetus and Placental Disk of a Dichorionic-Diamnionic Twin Gestation: A Case Report and Review of the Literature. Int J Gynecol Pathol. 2019;38(4):346-352. doi:10.1097/PGP.0000000000000506. PMID: 29620587.
101. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Neonatal severe primary hyperparathyroidism (MIM 239200). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 564–567.
102. Scott-Warren VL, Walker R. Perioperative management of patients with Mucolipidosis II and III: Lessons from a case series. Paediatr Anaesth. 2021;31(3):260-267. doi:10.1111/pan.14090. PMID: 33247997.
103. Pispati A, Pandey V, Patel R. Oral bisphosphonate-induced recurrent osteonecrosis of jaw with atypical femoral fracture and subsequent mandible fracture in the same patient: a case report. J Orthop Case Rep. 2018;8(3):85–8. doi:10.13107/jocr.2250-0685.1124. PMID: 30584526. PMC6298721.
104. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Dysostosis multiplex. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 109–116.
105. Langereis EJ, Wagemans T, Kulik W, Lefeber DJ, van Lenthe H, Oussoren E, et al. A Multiplex Assay for the Diagnosis of Mucopolysaccharidoses and Mucolipidoses. PLoS One. 2015;10(9):e0138622. doi:10.1371/journal.pone.0138622. PMID: 26406883. PMC4583541.
106. Ferreira CR, Devaney JM, Hofherr SE, Pollard LM, Cusmano-Ozog K. Hereditary fructose intolerance mimicking a biochemical phenotype of mucolipidosis: A review of the literature of secondary causes of lysosomal enzyme activity elevation in serum. Am J Med Genet A. 2017;173(2):501-509. doi:10.1002/ajmg.a.38023. PMID: 27797444. PMC10506159.
107. Velho RV, Alegra T, Sperb F, Ludwig NF, Saraiva-Pereira ML, Matte U, et al. A de novo or germline mutation in a family with Mucolipidosis III gamma: Implications for molecular diagnosis and genetic counseling. Mol Genet Metab Rep. 2014;1:98-102. doi:10.1016/j.ymgmr.2014.01.002. PMID: 27896079. PMC5121289.
108. Ceroni JRM, Spolador GM, Bermeo DS, Honjo RS, de Oliveira LAN, Bertola DR, et al. Clinical and radiological findings in Brazilian patients with mucolipidosis types II/III. Skeletal Radiol. 2019;48(8):1201-1207. doi:10.1007/s00256-019-3159-x. PMID: 30712120.
109. Hagemeijer MC, van den Bosch JC, Bongaerts M, Jacobs EH, van den Hout JMP, Oussoren E, et al. Analysis of urinary oligosaccharide excretion patterns by UHPLC/HRAM mass spectrometry for screening of lysosomal storage disorders. J Inherit Metab Dis. 2023;46(2):206-219. doi:10.1002/jimd.12597. PMID: 36752951.
110. Lai LM, Lachman RS. Early characteristic radiographic changes in mucolipidosis II. Pediatr Radiol. 2016;46:1713–20. doi:10.1007/s00247-016-3673-0. PMID: 27525427.
111. Burgac E, Kaplan İ, Köseci B, et al. Early diagnostic clues of mucolipidosis type II: significance of radiological findings. Am J Med Genet A. 2024;194:e63545. doi:10.1002/ajmg.a.63545. PMID: 38264826.
112. Hassan O, Lodha A, Lafontaine M, Lodha A, Al Awad E. Unlocking the Enigma: Investigating I-Cell Disease in a Newborn Through Placental Pathology. Indian J Pediatr. 2024;91(12):1278-1280. doi:10.1007/s12098-024-05073-6. PMID: 38376646.
113. Sheth J, Mistri M, Shah K, Chaudhary M, Godbole K, Sheth F. Lysosomal Storage Disorders in Nonimmune Hydrops Fetalis (NIHF): An Indian Experience. JIMD Rep. 2017;35:47-52. doi:10.1007/8904_2016_24. PMID: 27928775. PMC5585095.
114. Firth HV, Hurst JA. Skeletal dysplasia charts. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 853–855.
115. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Multiple sulfatase deficiency. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 750–760.
116. OMIM® Online Mendelian Inheritance in Man. #239200 HYPERPARATHYROIDISM, NEONATAL SEVERE; NSHPT. Johns Hopkins University; last edited 12/02/2026. https://omim.org/entry/239200. Accessed 28 October 2026.
117. Arowolo AT, Adeola HA, Khumalo NP. "Next generation sequencing identifies mutations in GNPTG gene as a cause of familial form of scleroderma-like disease". Pediatr Rheumatol Online J. 2017;15(1):88. doi:10.1186/s12969-017-0215-8. PMID: 29287597. PMC5747945.
118. Lewis JR, Gibson PH. Bilateral hip replacement in three patients with lysosomal storage disease: mucopolysaccharidosis type IV and mucolipidosis type III. J Bone Joint Surg Br. 2010;92:289–92. doi:10.1302/0301-620X.92B2.23104. PMID: 20130326.
119. Oussoren E, Wagenmakers MAEM, Link B, van der Meijden JC, Pijnappel WWMP, Ruijter GJG, et al. Hip disease in Mucopolysaccharidoses and Mucolipidoses: A review of mechanisms, interventions and future perspectives. Bone. 2021;143:115729. doi:10.1016/j.bone.2020.115729. PMID: 33130340.
120. Liang H, Liang Y, Zhan X, Chen Y, Yi M, Zhang H. Zoledronic acid ameliorates multiple organ pathologies in a murine model of mucolipidosis II by partially restoring mannose-6-phosphate-mediated lysosomal hydrolase activities. Mol Genet Metab. 2026;148(4):110191. doi:10.1016/j.ymgme.2026.110191. PMID: 42320387.
121. Simm PJ, Biggin A, Zacharin MR, Rodda CP, Tham E, Siafarikas A, et al. Consensus guidelines on the use of bisphosphonate therapy in children and adolescents. J Paediatr Child Health. 2018;54(3):223-233. doi:10.1111/jpc.13768. PMID: 29504223.
122. DeVile C, Allgrove J, Crowe B, Heathfield M. Drug monograph for osteogenesis imperfecta: use of intravenous pamidronate disodium. Protocol version 4 (DTC ref. IN00055). London: Great Ormond Street Hospital for Children NHS Foundation Trust; May 2011 (reviewed July 2016 and July 2021). Available from: https://web.archive.org/web/2024id_/https://media.gosh.nhs.uk/documents/Pamidronate_Protocol_-_Version_4_IN00055.pdf (accessed 2026 Oct 27).
123. Rauch F, Travers R, Plotkin H, Glorieux FH. The effects of intravenous pamidronate on the bone tissue of children and adolescents with osteogenesis imperfecta. J Clin Invest. 2002;110(9):1293-9. doi:10.1172/JCI15952. PMID: 12417568. PMC151613.
124. Laboratórios Pfizer, Lda. Pamidran 3 mg/ml e 9 mg/ml concentrado para solução para perfusão — Resumo das Características do Medicamento. INFARMED/Infomed, med_id 33494. Available from: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=33494&tipo_doc=rcm
125. Hospira, Inc. Pamidronate disodium injection [label]. ANDA075841. DailyMed set id 9b6a9a3b-71c9-4dcb-b857-5d1de25c3a23; effective version 2026-05-14. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=9b6a9a3b-71c9-4dcb-b857-5d1de25c3a23
126. European Medicines Agency. Aclasta (zoledronic acid): EPAR — product information. Available from: https://www.ema.europa.eu/en/documents/product-information/aclasta-epar-product-information_en.pdf
127. Sandoz Inc. Reclast (zoledronic acid) injection [label]. NDA021817. DailyMed set id 5a9b3737-9ce6-4a89-b76e-6aab79eba9cf; effective version 2026-02-03. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=5a9b3737-9ce6-4a89-b76e-6aab79eba9cf
128. Organon LLC. Fosamax Plus D (alendronate sodium/cholecalciferol) tablets [label]. NDA021762. DailyMed set id 10307e7e-9a84-4aa1-8c5c-4b209cffe4d1; effective version 2026-02-03. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=10307e7e-9a84-4aa1-8c5c-4b209cffe4d1
129. Wexler M. FDA grants orphan drug status to M6P gene therapy. Gaucher Disease News. 2021 Feb 8. Available from: https://gaucherdiseasenews.com/2021/02/08/fda-grants-orphan-drug-designation-gaucher-gene-therapy-m6p-therapeutics/ (accessed 2026 Oct 27).
130. Agência Nacional de Vigilância Sanitária (ANVISA). Dados abertos: medicamentos registados no Brasil (DADOS_ABERTOS_MEDICAMENTOS.csv). Available from: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv (accessed 2026 Oct 27).
131. Grewal S, Shapiro E, Braunlin E, Charnas L, Krivit W, Orchard P, et al. Continued neurocognitive development and prevention of cardiopulmonary complications after successful BMT for I-cell disease: a long-term follow-up report. Bone Marrow Transplant. 2003;32(9):957-60. doi:10.1038/sj.bmt.1704249. PMID: 14561999.
132. He SJ, Li DJ, Lv WQ, Tang WH, Sun SW, Zhu YP, et al. Outcomes after HSCT for mucolipidosis II (I-cell disease) caused by novel compound heterozygous GNPTAB mutations. Front Pediatr. 2023;11:1199489. doi:10.3389/fped.2023.1199489. PMID: 37484777. PMC10359890.
133. Shibazaki T, Hirabayashi K, Saito S, Shigemura T, Nakazawa Y, Sakashita K, et al. Clinical and laboratory outcomes after umbilical cord blood transplantation in a patient with mucolipidosis II alpha/beta. Am J Med Genet A. 2016;170A(5):1278-82. doi:10.1002/ajmg.a.37563. PMID: 26789537.
134. Naumchik BM, Gupta A, Flanagan-Steet H, Steet RA, Cathey SS, Orchard PJ, et al. The Role of Hematopoietic Cell Transplant in the Glycoprotein Diseases. Cells. 2020;9(6). doi:10.3390/cells9061411. PMID: 32517081. PMC7348849.
135. Wilson WR, Gewitz M, Lockhart PB, Bolger AF, DeSimone DC, Kazi DS, et al. Prevention of Viridans Group Streptococcal Infective Endocarditis: A Scientific Statement From the American Heart Association. Circulation. 2021;143(20):e963-e978. doi:10.1161/CIR.0000000000000969. PMID: 33853363.
136. Delgado V, Ajmone Marsan N, de Waha S, Bonaros N, Brida M, Burri H, et al. 2023 ESC Guidelines for the management of endocarditis. Eur Heart J. 2023;44(39):3948-4042. doi:10.1093/eurheartj/ehad193.
137. Matos L, Vilela R, Rocha M, Santos JI, Coutinho MF, Gaspar P, et al. Development of an Antisense Oligonucleotide-Mediated Exon Skipping Therapeutic Strategy for Mucolipidosis II: Validation at RNA Level. Hum Gene Ther. 2020;31(13-14):775-783. doi:10.1089/hum.2020.034. PMID: 32283951.
138. Gonçalves M, Encarnação M, Moreira L, Gaspar P, Duarte AJ, Coutinho MF, et al. Relevance of functional studies for assessing an antisense oligonucleotide-mediated exon skipping therapeutic strategy for mucolipidosis type II. Orphanet J Rare Dis. 2026;21(1). doi:10.1186/s13023-026-04430-2. PMID: 42310785. PMC13528052.
139. Doray B, Le S, Dickson PI. Letter: Exon 19 of the GNPTAB Gene is Indispensable for the Catalytic Activity of the Enzyme. Hum Gene Ther. 2026;37(17-18):815-818. doi:10.1177/10430342261468988. PMID: 42427214.
140. Fujita T, Terawaki S, Moriwaki T, Otomo T. M6PR/IGF2R double-knockout supplies therapeutic sources for mucolipidosis II and Niemann-Pick C2. J Biol Chem. 2026;302(10):113511. doi:10.1016/j.jbc.2026.113511. PMID: 42665096. PMC13631582.
141. Pyne NK, Bagel J, Shyng C, Odonnell P, Miyadera K, Srnak J, et al. Age-sensitive response of systemic AAV-mediated gene therapy in a newly characterized feline model of mucolipidosis II. Mol Ther. 2025;33(8):3808-3821. doi:10.1016/j.ymthe.2025.04.030. PMID: 40285357. PMC12461657.
142. Ko AR, Jin DK, Cho SY, Park SW, Przybylska M, Yew NS, et al. AAV8-mediated expression of N-acetylglucosamine-1-phosphate transferase attenuates bone loss in a mouse model of mucolipidosis II. Mol Genet Metab. 2016;117(4):447-55. doi:10.1016/j.ymgme.2016.02.001. PMID: 26857995.
143. Patient Worthy. ASGCT: datos de prueba de concepto compartidos en M002 para ML II [news item]. 10 jun 2021. Available from: https://patientworthy.com/es/?p=1297850 (accessed 2026 Oct 27).
144. Casazza K, Kimonis V, Whitley CB, Jarnes JR. Glycoproteinoses: Clinical features, therapeutic landscape, and regulatory pathways in rare lysosomal disorders. Mol Genet Metab. 2026;148(2):110129. doi:10.1016/j.ymgme.2026.110129. PMID: 42008923.
145. Poorthuis BJ, Wevers RA, Kleijer WJ, et al. The frequency of lysosomal storage diseases in The Netherlands. Hum Genet. 1999;105:151–6. doi:10.1007/s004399900075. PMID: 10480370.
146. Poupetová H, Ledvinová J, Berná L, et al. The birth prevalence of lysosomal storage disorders in the Czech Republic: comparison with data in different populations. J Inherit Metab Dis. 2010;33:387–96. doi:10.1007/s10545-010-9093-7. PMID: 20490927.
147. Okada S, Owada M, Sakiyama T, et al. I-cell disease: clinical studies of 21 Japanese cases. Clin Genet. 1985;28:207–15. doi:10.1111/j.1399-0004.1985.tb00388.x. PMID: 2998652.
148. Omar A, Abdul Rahman S, Mohamed R, et al. Prevalence of lysosomal storage disease (LSD) in Malaysia. Malays J Pathol. 2025;47:273–85. PMID: 40883240.
149. ISMRD — The International Advocate for Glycoprotein Storage Diseases. Mucolipidosis II; Mucolipidosis III gamma. Available from: https://ismrd.org/mucolipidosis-ii/