Wolfram Syndrome (WFS1- and CISD2-Related Disorders)

Wolfram syndrome type 1 (DIDMOAD), autosomal dominant WFS1 disorders and Wolfram syndrome type 2

Literature and regulatory information updated: 26 October 2026.

0. Initial Reference Table

Entities in scope

The OMIM numbers of the entities, taken from the tables of GeneReviews and of the Italian consensus, match those of the OMIM entries consulted. The WFS1 gene entry (606201) associates four phenotypes with it, all at 4p16.1: Wolfram syndrome type 1 (222300, AR), Wolfram-like syndrome (614296, AD), DFNA6/14/38 deafness (600965, AD) and cataract 41 (116400, AD); the WS2 entry (604928) places CISD2 (611507) at 4q24, with AR inheritance [1, 2, 5–7, 24–27]. Cataract 41 is flagged as a provisional gene–phenotype relationship, because it rests on a single family [27]. The dominant neonatal form has no OMIM number of its own and does not appear in the gene's phenotype table [2, 5]. The number 125853 does not designate a disorder specific to WFS1: it is the OMIM entry for type 2 diabetes mellitus (T2D), which brings together polygenic susceptibility and several monogenic forms, with 27 genes or loci listed; in the version edited on 21 April 2025, the phenotype–gene relationship table of that entry does not include WFS1, and the text mentions the gene only in relation to the association of common polymorphisms with the risk of type 2 diabetes [28]. The differential diagnosis includes monogenic diabetes and MODY, monogenic neonatal diabetes, hereditary optic atrophies, non-syndromic genetic deafness, hereditary congenital cataracts, mitochondrial disorders, HNF1B-related disease and Leber hereditary optic neuropathy.

1. Introduction

1.1. Original description and evolution of the concept

The syndrome was described in 1938 by Wolfram and Wagener, in the Mayo Clinic Proceedings, in four of eight siblings who combined juvenile diabetes mellitus and optic atrophy [2]. Over the following decades central diabetes insipidus and sensorineural hearing loss were added, and the acronym DIDMOAD (diabetes insipidus, diabetes mellitus, optic atrophy, deafness) came to designate the tetrad; some authors proposed DIDMOADUD to include urinary dysfunction, which is frequent and accounts for part of the morbidity [2]. The reference clinical description is the British national study by Barrett, Bundey and Macleod, published in 1995, which brought together 45 patients from 29 families before the molecular era and established the typical sequence of manifestations, their frequency and the neurodegenerative nature of the disease [16]. That study showed that diabetes and optic atrophy usually appear before 16 years, that 64% of patients have sensorineural hearing loss by 20 years and that 60% present, at a mean age of 16 years (range 5 to 32), at least one of six manifestations: cerebellar ataxia, peripheral neuropathy, intellectual disability, dementia, psychiatric illness or urinary tract atony [1, 16].

1.2. Molecular milestones

In 1998, two groups independently identified the WFS1 gene at 4p16.1: Inoue and colleagues described a gene encoding a transmembrane protein mutated in patients with diabetes mellitus and optic atrophy, and Strom and colleagues named the predicted protein “wolframin” and described the variants in patients with DIDMOAD [29, 30]. In 1999, analysis of 19 British, Pakistani and mixed Arab and African families identified WFS1 variants in 17 of the 19 probands and showed the dispersion of variants along the gene [31]. In 2007, Amr and colleagues identified, in Jordanian families of Palestinian origin, a homozygous variant in a gene encoding a zinc finger protein, ERIS, then designated the cause of Wolfram syndrome type 2; the gene is now known as CISD2 [32]. From 2006 the dominant forms of WFS1 were described: optic atrophy with deafness and impaired glucose regulation in a Danish family, the p.(Ala684Val) variant as a frequent cause of dominant optic atrophy with deafness, dominant adult-onset diabetes in a Finnish family, dominant congenital nuclear cataract in an Irish family and, in 2017, the syndrome of neonatal or infancy-onset diabetes with congenital deafness and congenital cataracts caused by de novo dominantly acting variants [17, 33–36].

Before cloning, recessive inheritance was inferred from sibships with several affected members and from the frequency of consanguineous parents in the series published between the 1940s and 1970s, and the demonstration of vasopressin deficiency in a child, in 1986, confirmed the central origin of the diabetes insipidus [24]. In 1994, Polymeropoulos and colleagues mapped the locus to the short arm of chromosome 4 in 11 families (maximum lod score of 6.46 with marker D4S431); in 1996, the study of 12 British families by Collier and colleagues confirmed linkage to 4p and revealed locus heterogeneity, with one family with an atypical phenotype, unlinked to 4p, in which optic atrophy preceded diabetes by more than ten years [24]. In 2000, El-Shanti and colleagues excluded linkage to 4p16.1 in three of four consanguineous Jordanian families, with 16 affected members, and mapped a second recessive locus to 4q22–q24, the origin of the designation Wolfram syndrome type 2 [6, 24]. Dominant low-frequency hearing loss, described in a large North American family in 1968 and linked to 4p16.3 in 1995 (DFNA6), was later mapped to the same region in a Dutch family (DFNA14) and in a six-generation Canadian family (DFNA38); in 2001, the groups of Bespalova and Young showed that the three loci corresponded to heterozygous WFS1 variants [26].

1.3. Current nomenclature and classifications

Three classifications coexist and do not fully overlap. GeneReviews uses “WFS1 spectrum disorder” and distinguishes classic WFS1-SD, caused by two pathogenic variants and inherited in an autosomal recessive manner, from non-classic WFS1-SD, caused by a heterozygous variant, inherited or de novo, which includes DFNA6/14/38 low-frequency hearing loss and the forms with an isolated finding or several related findings [1]. The 2025 consensus of the Italian Society of Diabetology and the Italian Society of Paediatric Endocrinology and Diabetology (SID/SIEDP) separates “Wolfram syndrome” proper (WS1 due to WFS1 and WS2 due to CISD2, both recessive) from the “WFS1-related disorders”, which group Wolfram-like syndrome, the syndrome of neonatal or infancy-onset diabetes with deafness and cataracts, congenital cataract 41, DFNA6 deafness and WFS1-associated diabetes [2]. The international Delphi consensus, released in July 2026 as a preprint not yet peer reviewed, reserves the name “Wolfram syndrome” for the biallelic WFS1 or CISD2 forms and defines four subtypes under the term “Wolfram Spectrum Disorders”: (A) childhood-onset WFS1 Wolfram syndrome, with diabetes and optic atrophy before 18 years and biallelic variants; (B) adult-onset WFS1 Wolfram syndrome, with diabetes and optic atrophy at 18 years or later, slower progression and a predominance of missense variants; (C) WFS1-related syndromic disease, with early-onset profound deafness, with or without optic atrophy, and a dominant or de novo variant; (D) CISD2 Wolfram syndrome, with diabetes and optic atrophy before 18 years, gastrointestinal ulcers and a platelet abnormality [3]. That consensus leaves outside its scope the non-syndromic dominant forms (isolated deafness, isolated optic neuropathy, isolated cataracts) and the neonatal form with congenital deafness and hypotonia [3]. The authors of the Washington University registry use the designation “Hattersley-Urano syndrome” for the severe dominant form described by De Franco and colleagues [18].

1.4. Definition of the entities in this article

Wolfram syndrome type 1 is the multisystem neurodegenerative disease caused by biallelic pathogenic WFS1 variants, whose minimum criteria are juvenile-onset diabetes mellitus and optic atrophy [1, 37]. Wolfram syndrome type 2 is the form caused by biallelic CISD2 variants, clinically overlapping, with a tendency to peptic ulcer, bleeding and a platelet aggregation defect [1, 38]. The dominant WFS1 disorders are a group of phenotypes caused by a single variant, ranging from isolated low-frequency hearing loss, congenital nuclear cataract and adult-onset diabetes to the combination of optic atrophy and deafness (Wolfram-like syndrome) and the severe neonatal form [1, 2]. Neonatal diabetes caused by WFS1 is covered here in detail and sits within the group of monogenic neonatal diabetes, including Wolcott-Rallison syndrome.

2. Molecular Genetics and Pathophysiology

2.1. The WFS1 gene and wolframin

WFS1 is located at 4p16.1, has eight exons and encodes wolframin, a glycoprotein of 890 amino acids and about 100 kDa, with nine transmembrane domains, inserted in the endoplasmic reticulum membrane and sensitive to endoglycosidase H [1–3]. Expression is highest in the brain, the pancreas (β cells), the heart and the retinal ganglion cells and optic nerve glia [1, 39]. The exact function has not been established. The available data indicate that wolframin modulates filling of the endoplasmic reticulum calcium store, negatively regulates the reticulum stress response and stabilises subunits of V-ATPase (ATP6V1A) and Na⁺/K⁺-ATPase (ATP1B1) against degradation [39]. Its deficiency causes chronic endoplasmic reticulum stress, activation of the unfolded protein response (UPR), mitochondrial dysfunction, impaired cell cycle progression and death of β cells and neurons [1, 3]. The syndrome is therefore considered a prototype endoplasmic reticulum disease, and the cells most dependent on protein folding and calcium regulation (β cells, retinal ganglion cells, brainstem neurons) are those that fail first [1, 39, 40]. There is also in vitro evidence of an association with systemic inflammation [1]. In Saudubray, WFS1 appears among the defects of vesicular trafficking, with extraneurological manifestations listed for cardiomyopathy, cataract, retinopathy, renal dysplasia, diabetes and genital abnormalities [41].

The gene occupies, on GRCh38, the coordinates 4:6,269,850–6,303,265 [5]. The 3.6 to 3.7 kb transcript is detected in all human tissues studied, with expression in the pancreatic islets higher than in the exocrine pancreas; human wolframin shares 87% of its amino acid sequence with that of the mouse, is organised in the membrane in complexes of about 400 kDa and depends on N-glycosylation for its biogenesis and stability [5]. In Xenopus oocytes, the protein generated a cation conductance blocked by magnesium or calcium and increased by inositol trisphosphate, which led to the proposal that it acts as a reticulum calcium channel or as a regulator of such channels; in other models, Wfs1 expression increases with reticulum stress, under the control of the IRE1α and PERK pathways, and during glucose-stimulated insulin secretion, and its silencing in β cells causes reticulum stress and cellular dysfunction [5]. The C-terminal domain, facing the reticulum lumen, binds the β1 subunit of Na⁺/K⁺-ATPase, whose levels are reduced in patient fibroblasts [5]. In the missense variants of the recessive form, the documented mechanism is a reduction in protein dosage: the mutant proteins are unstable, have a short half-life and are removed by the proteasome, and in a patient compound heterozygous for a truncating and a missense variant, the messenger RNA, at half the level of controls, came only from the missense allele [5]. The Wfs1-null mouse develops glucose intolerance or diabetes due to insufficient insulin secretion, with a reduced calcium response, increased islet apoptosis under high glucose or reticulum stress inducers and progressive β-cell loss, whereas α cells, which barely express the protein, are preserved [5]. In neurons derived from patients' induced pluripotent stem cells, the lack of wolframin altered neurite outgrowth and the expression of axon guidance genes; restoration of the protein corrected the alteration, and valproate prevented it through a mechanism that did not depend on protection against reticulum stress, suggesting that an early defect of axon guidance contributes to neuronal loss [24].

2.2. Tissue pathophysiology

In diabetes, β-cell loss is progressive and non-autoimmune; residual insulin production persists in some patients for several years after diagnosis, which explains the prolonged remission phase and the lower insulin requirements than in type 1 diabetes [37, 42]. In optic atrophy, the lesion begins in the axons of the retinal ganglion cells and progresses to loss of the cell bodies, with axonal loss and demyelination of the optic nerves, chiasm and optic tracts [2]. Hypothalamic dysfunction explains arginine vasopressin deficiency and anterior pituitary hypofunction (growth hormone and corticotrophin deficiency), and magnetic resonance imaging shows loss of the posterior pituitary bright spot [2, 43]. In autopsy material from patients with AVP deficiency, there was loss of vasopressinergic neurons in the supraoptic nucleus and, in the paraventricular nucleus, cells with the vasopressin precursor but without the processed hormone, in the absence of the convertase PC2 and the chaperone 7B2; the authors interpret this as a defect of precursor processing added to neuronal loss [24]. In the central nervous system, imaging studies suggest two components: an early alteration of brain development (reduced intracranial volume, smaller brainstem, cerebellum and optic radiation) and later neurodegeneration [2]. Deafness may result from dysfunction at several points of the auditory pathway [2].

2.3. Dominant forms: a different mechanism

The variants of the classic recessive form are largely null variants leading to absence of the protein. The dominant variants described by De Franco and colleagues (p.(Glu809Lys), p.(Glu830Ala) and p.(His313Tyr)) form insoluble high-molecular-weight aggregates and induce endoplasmic reticulum stress under basal conditions and after stimulation with thapsigargin, to a greater degree than the recessive variant p.(Pro724Leu) used as a control [17]. Because the phenotype of these patients is more severe than that of complete homozygous loss, the mechanism is not a simple loss of function; it is a dominant effect that generates reticulum stress [2, 17]. The dominant variants associated with optic atrophy with deafness are missense and lie, in six of seven cases in the Washington University registry, in the luminal domain of the endoplasmic reticulum [18]. GeneReviews, updated in 2022, does not detail these mechanisms and states that no genotype–phenotype correlations have been identified, a position that later literature has corrected [1, 2].

Analysis of the variant spectrum supports this distinction. In the recessive form, the variants are distributed throughout the coding region and about 64% are inactivating; among the 10 variants then known in low-frequency hearing loss, none predicted premature termination and 9 clustered in the C-terminal domain, which led to dominant non-syndromic deafness being attributed to non-inactivating variants [5]. The same variant may, however, appear in different dominant phenotypes: p.(Glu864Lys) causes Wolfram-like syndrome in a Danish family and a French family and isolated low-frequency hearing loss in two Japanese families [5, 25].

2.4. The CISD2 gene and Wolfram syndrome type 2

CISD2 is located at 4q24 according to NCBI Gene and the international consensus; the Italian consensus places it at 4q22 [2–4]. It encodes ERIS (endoplasmic reticulum intermembrane small protein), also called NAF-1 or miner1, a protein associated with the endoplasmic reticulum and mitochondria that binds iron–sulphur clusters, regulates autophagy by antagonising BECN1, is required for BCL2-mediated depletion of reticulum calcium stores and participates in coupling between the reticulum and mitochondria [3, 38]. Wolframin and ERIS do not interact directly [1, 2]. The Cisd2 knockout mouse develops premature ageing with features of Wolfram syndrome [38]. In that model, the protein localises mainly to the outer mitochondrial membrane, and mitochondrial degradation and dysfunction, with autophagic cell death and a respiratory defect, precede nerve and muscle degeneration; the phenotype includes corneal opacities and degeneration, thinner bones and hair, reduced muscle mass and a shortened lifespan, which led the authors to consider WS2 a partly mitochondrially mediated disease [6]. The human variants recorded in OMIM are the homozygous variant of the Jordanian families, a homozygous intragenic deletion in an Italian adolescent girl born to unrelated parents and a homozygous splice variant in two Italian sisters without WFS1 variants, with heterozygous parents [6]. In fibroblasts from patients with WS2, mitochondrial labile iron increases by 25% and mitochondrial reactive oxygen species by 28% (p<0.0001), changes that deferiprone combined with N-acetylcysteine reversed in vitro [44].

2.5. Variant spectrum and nomenclature

The LOVD database listed 358 pathogenic or likely pathogenic WFS1 variants in March 2024, distributed along the coding sequence and concentrated in exon 8, the largest; most families have private variants [2]. The types include missense, nonsense, small insertions and deletions with or without frameshift and splice variants; three intragenic deletions of one or more exons and structural rearrangements have been described [1, 2, 45]. Few CISD2 variants are known, mainly deletions and missense variants [2]. Variants should be described on the MANE Select transcript NM_006005.3 for WFS1 and NM_001008388.5 for CISD2; the transcript NM_001145853 that appears in some databases is an alternative transcript [8]. The following table brings together the variants cited in this article, normalised in VariantValidator (GRCh38) and cross-checked against ClinVar.

Table sources: VariantValidator and ClinVar, consulted in October 2026 [8, 46], and the original articles [15, 17, 18, 32–36, 42, 44, 47–52]. The CISD2 variant c.109G>C illustrates the difference between the protein consequence predicted by an annotator, a missense change, and the effect demonstrated on the RNA, skipping of exon 2 [8, 32].

The allelic variants selected by OMIM include several published with historical numbering. Inoue and colleagues (1998) and Zalloua and colleagues (2008) counted nucleotides with an offset of 170 positions relative to numbering from the initiation codon: “2341C>T” corresponds to c.2171C>T, “2289G>A” to c.2119G>T and “2819delC” to c.2649del; the correspondence was confirmed by the dbSNP notation and by VariantValidator, and not by isolated calculation [5, 8]. The following table brings together the OMIM variants not included in the previous table, with the notation normalised in VariantValidator (GRCh38, NM_006005.3) or, where indicated, that of the dbSNP record given by OMIM.

Sources: OMIM, WFS1 gene entry [5], VariantValidator [8] and dbSNP. OMIM designates the variant c.683G>A as R228Q, but on the MANE transcript this change gives p.(Arg228His); OMIM also describes the variant of the WFS1(LIB) allele as a G>A transition, whereas the valine-to-phenylalanine change requires c.2119G>T [5, 8].

2.6. Genotype–phenotype correlations

The meta-analysis by López de Heredia and colleagues and the international Washington University registry showed three patterns: the greater the number of null variants, the earlier and more severe the disease; two nonsense or frameshift alleles are associated with earlier onset of diabetes and optic atrophy than zero or one allele of that type; and the number of in-frame variants in transmembrane domains has a dose effect on age at onset [2, 53, 54]. In 2026, Oiknine and colleagues proposed a genotype-based severity score, applied to 324 patients with documented ages at onset, with six classes defined by variant type (in-frame or out-of-frame) and location relative to the transmembrane domains [48].

The score was associated with earlier age at onset of diabetes (except in class 3), with significant differences between classes 2 and 3 (p=0.008) and between 4 and 5 (p=0.0015); the association with optic atrophy was less consistent and there was no correlation with deafness or AVP deficiency [48]. Grouped into three levels defined after the analysis, the median ages at onset of diabetes were 9.0, 6.5 and 4.6 years in the mild, moderate and severe levels [48]. The study is exploratory, did not use time-to-event analysis and the tool is not validated for individual prediction [48]. The p.(Arg558Cys) variant, in-frame and outside the transmembrane domains, is associated with a mild form, frequent in Ashkenazi Jews, and is the example cited by the international consensus for the adult-onset subtype [3, 48]. Variability is great even between siblings with the same variants, and the individual course cannot be predicted from the genotype alone [1]. In the dominant forms, the three variants of De Franco and colleagues included the two most severe patients, both with p.(Glu809Lys), suggesting a correlation within this group [17].

The OMIM entries summarise the studies on which these correlations were based. In a meta-analysis of 96 patients from five studies, Cano and colleagues (2007) associated two inactivating variants with earlier onset of diabetes and optic atrophy and more complete clinical expression [24]. In the French series of 59 patients by Chaussenot and colleagues (2011), missense variants, in one or two copies, significantly delayed the onset of diabetes and optic atrophy, but neither the appearance nor the age at onset of neurological signs correlated with genotype [24]. López de Heredia and colleagues grouped 178 variants, concentrated in the transmembrane domains, the glycosylation sites, between amino acids 94 and 237 and in the last 100 amino acids (only 6 present in more than 5% of patients), into three classes of predicted effect (A, no protein; B, reduced expression of defective protein; C, expression of defective protein), and the genotypic classes of 337 patients correlated with age at onset of diabetes, AVP deficiency and deafness [24]. In contrast, Hardy and colleagues (1999), with 19 British families, found no clear correlation between variant and severity, and the only association of a distinct phenotype with a specific variant is the central respiratory failure described in homozygotes for c.2648_2651del, absent in five compound heterozygotes for the same variant [5, 24].

2.7. Heterozygotes, penetrance and allelic disorders

The same gene produces recessive and dominant disease, and some variants appear in the compound heterozygous state in patients with the classic form and in the simple heterozygous state in patients with Wolfram-like syndrome [2]. It is not clear whether the two situations are distinct entities or the same recessive condition with reduced penetrance in heterozygotes, which complicates counselling of carriers [2]. Lusk and colleagues documented the segregation of two variants in a family with coexisting dominant and recessive forms [9]. GeneReviews found no conclusive evidence of an increased risk of spectrum components in the heterozygotes (parents or siblings) of patients with biallelic variants, whereas the Italian consensus stresses that heterozygous carriers are not necessarily healthy [1, 2]. The international consensus recommends excluding other causes, such as autoimmune type 1 diabetes, before attributing a symptom in a carrier to WFS1, and notes that only one known dominant variant is associated with isolated diabetes; in most cases, dominant WFS1 variants do not cause monogenic diabetes [3]. GeneReviews also includes susceptibility to type 2 diabetes (OMIM 125853) among the OMIM entries of the gene [1]. The corresponding OMIM entry describes the gene-centred association study by Sandhu and colleagues (2007), which identified, in several large cohorts, two intronic WFS1 polymorphisms, rs10010131 and rs6446482, associated with the risk of type 2 diabetes (p = 1.4 × 10⁻⁷ and p = 3.4 × 10⁻⁷ in the pooled set); the risk allele was the major allele in both cases, with a frequency of 60%, and neither has an evident biological function [28]. In the gene entry, OMIM reclassified the two variants in 2026 as polymorphisms, based on allele frequencies of about 0.62 and 0.63 in gnomAD v4.1 [5]. In a Spanish series of 23 patients with sensorineural hearing loss and type 2 diabetes, three had heterozygous WFS1 missense variants of unclear contribution; one of them, p.(Val871Met), was present in a father without deafness and did not segregate with the disease in another family [5, 25]. This is a population susceptibility effect, distinct from the rare strong-effect variants of the Mendelian forms of the gene, and has no utility in individual diagnosis [28]. Very rarely, WFS1 variants are associated with recessively inherited isolated optic atrophy, and an observational study at Hôpital Necker begun in February 2026 (NCT07336966) seeks to clarify whether this form exists [1, 55].

3. Clinical Spectrum and Manifestations by System

3.1. Data sources and what each one measures

The natural history of the classic form rests on few cohorts, almost all from specialist centres, and only a minority of the published cases underwent extensive assessment of all organs [1]. The 1995 British cohort (45 patients, 29 families, pre-molecular era) provides the reference frequencies and ages; the EURO-WABB registry (121 participants with a genetically confirmed diagnosis) gives the frequencies used in the European criteria; the international Washington University registry, in St. Louis, underpins the recent genotype, imaging and hearing studies; and the national cohorts (France, 96 patients; Italy; Spain; Germany; Ecuador; Portugal) add regional data [15, 16, 37, 45, 52, 54, 56]. The EURO-WABB registry frequencies are proportions in a group of referred patients of heterogeneous ages; they are not lifetime risks. The following table summarises, from GeneReviews, which manifestations are common, uncommon or increase with age.

Source: GeneReviews, Table 2 [1].

Three series described in the OMIM entries complete these data. In the French series of Chaussenot and colleagues (2011; 59 patients from 49 families), diabetes appeared in 97%, at a median of 6 years, and optic atrophy in 98.5%, at a median of 10 years; there were neurological complications in 53%, at a median of 15 years, an age lower than in earlier series, in the form of brainstem or cerebellar dysfunction (55% of patients with neurological signs), peripheral neuropathy (39%) and cognitive impairment (32%); urinary tract abnormalities appeared in 53% (median 12 years), deafness in 46% (16 years), AVP deficiency in 29% (15 years), gastrointestinal abnormalities in 12% and bilateral cataracts in 5%; there were psychiatric symptoms in 23 patients, a figure presented as 59% with a denominator the entry does not make explicit, and primary gonadal atrophy in 8 of 32 men [24]. The authors stressed that death results mainly from respiratory failure or dysphagia due to brainstem involvement [24]. In the Lebanese series of Medlej and colleagues (2004; 31 patients from 17 families, clinical diagnosis based on insulin-dependent diabetes and optic atrophy), AVP deficiency affected 87% and audiogram-confirmed deafness 64.5%, with heart disease and anterior pituitary dysfunction among the new findings; the methods of the time identified WFS1 variants in only 3 of the 17 families (23.5%), without mitochondrial DNA alterations [24]. The review of 412 patients published over the previous 15 years by López de Heredia and colleagues (2013) showed that 15% did not meet the criterion of juvenile diabetes and optic atrophy and that these two manifestations are not always the first; among 392 patients with documented ages, diabetes appeared in the first decade, optic atrophy at the start of the second, AVP deficiency and deafness in the second and urological and neurological abnormalities in the third, with urological symptoms at a median of 20 years in 76 patients and peaks at 13, 21 and 33 years [24]. These are series of referred or published patients, subject to selection and publication bias, and their proportions are not lifetime risks.

3.2. Diabetes mellitus

Diabetes is usually the first manifestation. In the British cohort, the median age at onset was under 10 years (range from under 1 to 17 years); the international consensus gives a median of 6 years and ISPAD (International Society for Pediatric and Adolescent Diabetes) a mean age of about 6 years [1, 3, 16, 57]. In the EURO-WABB registry, 87% had diabetes before 16 years and 4% after that age [37]. Almost all patients eventually need insulin [2]. The diabetes is non-autoimmune, with type 1 diabetes autoantibodies usually absent, although adults with anti-β-cell autoimmunity have been described and the Ecuadorian project found autoantibodies in 3 of 26 patients [2, 15]. Ketoacidosis at presentation is rare and there is often a prolonged remission phase [37]. Compared with type 1 diabetes, it has earlier onset, lower glycated haemoglobin (HbA1c), lower insulin requirements and fewer microvascular complications: in the cohort of Cano and colleagues, HbA1c was 7.72 ± 0.21% against 8.99 ± 0.25% (p=0.002) and the insulin dose 0.71 ± 0.07 against 0.88 ± 0.04 U/kg/day (p=0.0325), with measurable C-peptide more than 8 years after diagnosis in some patients [42]. A higher HbA1c was associated with faster progression of neurodegenerative and endocrine symptoms [42]. Severe hypoglycaemia is frequent, and hypoglycaemia awareness may be severely impaired [1, 37, 58].

The 2019 French PNDS places the mean onset of diabetes at about 7 years and detection of optic neuropathy at 9 or 10 years, noting that the disease may begin in the optic nerve [23]. It attributes the rarity of ketoacidosis to residual insulin secretion, which persists for many years despite the constant loss of β cells, and describes the typical complications of diabetes as infrequent, with diabetic retinopathy in 7% and diabetic neuropathy or nephropathy in 9%, proportions taken from published series and not from a population registry; this rarity does not obviate monitoring of microalbuminuria and creatinine [23].

In Lebanon, systematic testing of WFS1 in 399 probands with juvenile-onset insulin-dependent diabetes found biallelic variants in 22 (5.5%): 17 with Wolfram syndrome and 5 with isolated non-autoimmune diabetes [5, 24]. Among the 38 affected probands and relatives, 11 (29%) had isolated diabetes, all carriers of the founder complex allele WFS1(LIB); two patients initially classified as having isolated diabetes developed optic atrophy during the study, and the oldest with isolated diabetes were 20 and 23 years old, so the follow-up does not allow the conclusion that extrapancreatic manifestations will not appear [5, 24].

3.3. Optic atrophy and other ocular manifestations

Optic atrophy eventually occurs in all patients with the classic form, with a median onset before 10 years and progression over years to severe visual impairment; in the EURO-WABB registry, 80% had optic atrophy before 16 years and 7% after [1, 37]. The loss involves colour and peripheral vision; an acuity of 6/60 corresponds to the threshold for registered visual impairment in the United Kingdom and legal blindness in the United States, and most progress to 3/60 (about 20/400) [1, 2]. Visual loss appears about four years after diabetes and may be confused with diabetic retinopathy [57]. Other ophthalmological findings include diabetic retinopathy, retinal abnormalities, cataracts, glaucoma, nystagmus and abnormal pupillary reflexes [2]. In the cohort from the Centro Hospitalar e Universitário de São João (Porto), with 11 patients from 6 families assessed between April and November 2019, optic atrophy was present in all, corrected acuity ranged from light perception to 0.1 logMAR (mean 0.97 ± 0.69; 10 of 22 eyes at 1 logMAR or worse) and optical coherence tomography angiography (OCT-A) showed reduced peripapillary, intradiscal and parafoveal superficial capillary plexus vessel density compared with 18 controls (p<0.001) [52]. The authors interpret the microvascular change as a possible contributor to neurodegeneration [52]. In the full text of the same study, the patients were 26.6 ± 9.8 years old (11 to 40 years) and the controls, matched for age and sex, 33.0 ± 7.9 years [52]. Diabetes, optic atrophy and sensorineural hearing loss appeared on average at 10.0, 14.6 and 14.0 years, followed by neurological (19.9 years), psychiatric (22.7 years) and urological (23.1 years) abnormalities; the only patient with AVP deficiency had diabetes and optic atrophy diagnosed at 10 years [52]. Six eyes of 3 patients (27.3%) had posterior subcapsular opacity or cataract, already operated on before the study; mean intraocular pressure was 13.7 mmHg, 2 patients had conjugate horizontal jerk nystagmus, 1 had esotropia, and no eye had diabetic retinopathy [52]. On structural OCT, mean retinal nerve fibre layer (RNFL) thickness was 52.3 µm against 90.3 µm in controls (adjusted difference −39.3 µm; 95% CI −45.7 to −32.9) and that of the macular ganglion cell complex (GCC) 54.3 µm against 80.6 µm (−26.7 µm; 95% CI −32.0 to −21.4) [52]. On OCT-A, the vessel density of the whole optic disc image was 39.4% against 47.8% (−9.3 points; 95% CI −11.1 to −7.5); the reduction affected all peripapillary sectors and was greatest in the temporal sector (−12.9 points), which the authors relate to preferential involvement of the papillomacular bundle [52]. In the parafoveal superficial capillary plexus, density was 40.1% against 44.8% (−5.4 points; 95% CI −7.4 to −3.4), with a reduction in the nasal, inferior and superior sectors and no difference in the temporal sector (p = 0.961) [52]. Patients with acuity of 1 logMAR or worse had thinner RNFL and GCC than the others (−14.6 µm, p = 0.001, and −12.6 µm, p<0.001), but not different vessel densities, so the study cannot determine whether vascular loss is primary or secondary to neuronal loss [52]. The cross-sectional design, the small sample, the inclusion of both eyes, a single OCT-A device and the inability to fixate in 2 patients limit the conclusions; the article also mentions 12 patients in the discussion and a “case 12” in the figures, a number that does not match the 11 patients in the table [52].

The PNDS describes as the most usual ophthalmological presentation a bilateral fall in acuity at about 9 years [23]. Dyschromatopsia, on a red–green confusion axis, may precede the loss of acuity and should, on its own, prompt a search for optic neuropathy; photophobia, when present, is mild [23]. Cataract, of pulverulent or subcapsular type and sometimes congenital, may appear before diabetes and optic neuropathy and add to fibre loss as a cause of low vision, with an indication for phacoemulsification; cases of congenital glaucoma and, rarely, moderate macular pigmentary changes with a normal electroretinogram, similar to those of MIDD, have been described [23]. With progression, difficulty extends from pale to bright colours, peripheral scotomas disturb walking and nystagmus may appear when acuity falls to 1/10 or less [23].

3.4. Hearing

In the classic form, sensorineural hearing loss affects about 66% of patients, with a median onset of 12.5 years; it ranges from congenital deafness to mild losses beginning in adolescence, preferentially affects the high frequencies, progresses slowly and does not differ between the sexes [1, 2, 16]. In the EURO-WABB registry, 48% had hearing loss [37]. The longitudinal study of 40 patients from the St. Louis registry showed worsening of high-frequency thresholds and of the speech intelligibility index over time, but a subclinical annual change [59]. In the dominant forms, the pattern is different: in DFNA6/14/38 deafness, the hearing loss is congenital, slowly progressive and affects frequencies below 2,000 Hz, often diagnosed only after language acquisition, and the decline in speech recognition follows the degree of hearing loss [1]. Low-frequency hearing loss is unusual among dominant forms of deafness, which are generally post-lingual and high-frequency; WFS1, DIAPH1 and COL11A2 are the main exceptions [60, 61]. Why the classic form affects the high frequencies and DFNA6/14/38 the low frequencies is unknown [1].

In the classic form, a small fraction of patients have congenital deafness [24]. In DFNA6/14/38 deafness, frequencies of 2,000 Hz and below are the most affected, tinnitus is frequent and there is no vertigo; because the high frequencies are preserved, speech comprehension is maintained, although presbycusis or noise may add high-frequency loss later [26]. The hearing loss worsens without reaching profound deafness, unlike low-frequency hearing loss due to DIAPH1 (DFNA1), which progresses to profound deafness by the fourth decade [26]. Age at onset varies between families: in the Dutch DFNA14 family, loss generally began in the second decade, and the Japanese family with p.(Lys634Thr) had no tinnitus [5, 26]. In Japan, WFS1 sequencing in 206 probands with dominant non-syndromic deafness and 64 with recessive or sporadic deafness identified variants in three families, all with dominant low-frequency hearing loss and with variants already described in Europeans, p.(Ala716Thr) and p.(Glu864Lys), suggesting recurrent mutational sites [26]. In a North American family with p.(Arg859Gln), two affected women had Crohn's disease and Graves' disease, and the authors hypothesised a relationship with WFS1, without demonstration [26].

In the series cited by the PNDS, deafness affects 62 to 72% of patients, with a mean age at onset between 10 and 16 years; it is usually post-lingual, which allows normal development of oral language, although onsets before 2 years have been described [23]. The degree ranges from moderate to profound, the lesion is endocochlear and there is loss of cochlear nerve fibres and of neurons of the cochlear nucleus and inferior colliculus [23]. Balance disturbances appear at about 30 years, sometimes earlier, and are mainly due to ataxia, but vestibular tests also show abnormalities that may contribute to them [23]. Disorders of smell and taste of neurological origin and dysphagia with chewing difficulty and aspiration have been described [23].

3.5. Nervous system and mental health

In the British cohort, 62% had neurological abnormalities before molecular confirmation; GeneReviews gives a mean age of 30 years (range 5 to 44) and the Italian consensus a median onset of 15 years with the same range; there was truncal or gait ataxia in 15 of 45 and episodes of central apnoea in 5 of 45 [1, 2, 16]. In the EURO-WABB registry, 29% had neurological signs (ataxia, epilepsy, cognitive impairment) [37]. Current experience is of symptomatic findings by the fourth decade, with presymptomatic onset between the first and second decades, progressive and due to generalised brain atrophy with involvement of the brainstem and cranial nerves [1, 45]. The spectrum includes anosmia, hypogeusia, ataxia, nystagmus, gaze palsies, dysarthria, dysphagia, areflexia, epilepsy, headaches, central apnoea and dysautonomia (orthostatic hypotension, hyperpyrexia, hypothermia, constipation, gastroparesis) [2]. Seizures are uncommon and should prompt a search for another cause [3]. On magnetic resonance imaging of 30 patients followed for a median of 5 years, the abnormalities increased between the first and last visits, and they are also present in young patients with few symptoms [2, 43].

Source: Samara and colleagues, 2020, as cited by GeneReviews [1, 43].

According to the PNDS, neurological signs have a median age at onset of 15 years, with two peaks, and are rarer or absent with CISD2 variants; the cerebellar syndrome (unsteady gait, dysmetria, impaired coordination, dysarthria and, sometimes, nystagmus) is the most frequent involvement, at 54%, without an explicit denominator in the document [23]. The PNDS neurological table adds to the spectrum already described psychomotor delay or learning difficulties from childhood and, rarely, mild intellectual disability; dysautonomia with sweating abnormalities (hypohidrosis, hyperhidrosis or anhidrosis), gastroparesis, constipation and impaired thermoregulation; anosmia or hyposmia from the second or third decade; sleep disturbances (frequent awakenings, snoring, enuresis, hypersomnolence), often linked to the nocturnal voiding of neurogenic bladder and AVP deficiency; generalised seizures (myoclonic, tonic–clonic) and, more rarely, focal seizures, in the advanced stages; and late and insidious cognitive decline, with executive dysfunction and apraxia [23]. Among the psychiatric disorders, which it considers frequent, it includes oppositional disorder and eating disorders [23]. On magnetic resonance imaging, the earliest signs are cerebral, cerebellar and brainstem atrophy, especially of the ventral pons and the middle cerebellar peduncles; atrophy of the optic pathways, thinning of the hypothalamus and infundibulum and abnormalities of the substantia nigra may also be seen, and, in AVP deficiency, loss of the T1 posterior pituitary bright spot [23]. Kinetic ataxia, sensory deficits and neurogenic digestive and urinary disturbances add up to reduce independence in fine tasks (dressing, cooking, cutting) and walking [23].

The risk of psychiatric illness, including suicidal behaviour, is increased; the spectrum includes anxiety, panic attacks, depression, mood swings, sleep disturbances and psychosis, and dementia occurs in some patients as part of the neurodegeneration [1, 2]. In the EURO-WABB registry, 26% had a psychiatric disorder; intellectual disability is not a common feature [1, 37].

In a series of 68 homozygous patients studied by Swift and colleagues (1990), 60% had episodes of severe depression, psychosis or organic brain syndrome, with compulsive verbal and physical aggression [24]. Neurogenic respiratory failure, accompanied by startle myoclonus, Parinaud syndrome, axial rigidity and marked brainstem atrophy, was described in two pairs of siblings [24]. In patients homozygous for c.2648_2651del, central respiratory failure is the distinguishing feature: the patient described by Hardy and colleagues, with severe brainstem atrophy, required a tracheostomy and had a sister who died at 28 years from the same cause, and a woman described by Sam and colleagues, with diabetes at 10 years, neurogenic bladder at 13 and optic atrophy at 15, without deafness or AVP deficiency, had a respiratory arrest at 32 years and became tracheostomy-dependent [5, 24].

3.6. Arginine vasopressin deficiency and other endocrinopathies

Arginine vasopressin deficiency (the current designation for central diabetes insipidus) affects 72% of patients in the British cohort, with a median onset of 15.5 years; the age range may reflect diagnostic delay, because the polyuria is confused with that of poor glycaemic control and that of bladder dysfunction [1, 37]. In the EURO-WABB registry, the frequency was 42% [37]. Anterior pituitary hypofunction of hypothalamic origin can cause growth hormone and corticotrophin deficiency; hypothyroidism, central or non-autoimmune, has an unestablished frequency; most patients reach normal adult height [1, 2]. Hypogonadism is more frequent and more severe in men, hypo- or hypergonadotropic; in women, menstrual irregularities and delayed puberty are described [1, 2]. In a review of 21 men followed in Birmingham and Münster (16 to 30 years), puberty was severely delayed or arrested in 57%, requiring testosterone to complete it; 28.6% had compensated subclinical hypergonadotropic hypogonadism, 9.5% hypogonadotropic hypogonadism and 43% normal testicular function [62]. One man had oligozoospermia at 20 years and azoospermia at 25, and biopsy showed Leydig and Sertoli cell abnormalities, tubular atrophy and meiotic arrest; erectile dysfunction was reported by 8 of 9 men with a record (88.9%), even with adequate testosterone [62]. The frequencies of hypogonadism vary between 7.1% and 50% depending on the cohort [42].

The PNDS gives a frequency of AVP deficiency of about 42% (38 to 75% depending on the series), with onset in the second decade, after diabetes and optic atrophy and before ataxia and deafness, and considers it central, probably due to impaired synthesis of antidiuretic hormone [23]. For hypogonadism, it indicates 50% of adolescents, with equal involvement of both sexes: in boys, pubertal arrest between 12 and 16 years, incomplete virilisation, testes that stop growing, testosterone that does not rise at puberty and very high LH and FSH, with possible impairment of fertility; in girls, late menarche or irregular cycles [23]. Hypothyroidism, with fatigue, constipation, bradycardia and dry skin, may begin in adolescence [23].

3.7. Urinary tract, kidney and bowel

Urinary problems affect 60 to 90% of patients [2]. In the British cohort, 16 of 29 affected individuals had signs of neurogenic bladder (hydroureter, incontinence, recurrent infections), with a median onset of 22 years (10 to 44) [1]. Urodynamics shows incomplete emptying or complete bladder atony, with progression to megabladder and a risk of acute obstruction; large-capacity atonic bladders and low-capacity high-pressure bladders coexist [2, 63]. In the series of men from Birmingham and Münster, neurogenic bladder was present in 78.6% of the British men and in 6 of 7 German men [62]. Diagnostic delay leads to severe infections and acute and chronic kidney injury [2]. The Oxford Desk Reference mentions renal tract anomalies, such as hydronephrosis [64]. In the Ecuadorian cohort, a glomerular filtration rate below 60 mL/min occurred only in patients with WS1 (12%) [15]. Bowel dysmotility, with constipation, chronic diarrhoea and faecal incontinence, affects up to 25% of patients, and coeliac disease has been described [1, 2].

The PNDS reports urinary manifestations in about 90% of patients, with a median age at onset close to 15 years, earlier than in older series, which placed them in the third decade [23]. It describes stress incontinence, bladder overactivity, dysuria, detrusor overactivity or underactivity and detrusor–sphincter dyssynergia, and ureteral and pelvicalyceal dilatation that is rarely isolated; the impairment of renal function is mixed, due to diabetic nephropathy and to dilatation of the upper urinary tract [23]. Uronephrological complications affect 20% of patients, half of those with urinary symptoms report an impact on quality of life, and the polyuria of AVP deficiency worsens bladder symptoms, which justifies screening from childhood [23]. In the summary intended for the attending physician, the PNDS estimates that urinary abnormalities cause functional discomfort in one in three patients and recalls the proposal to extend the acronym to DIDMOADUA, because of the frequency of neurogenic bladder, hydroureter and hydronephrosis [23].

3.8. Other systems

GeneReviews lists structural congenital heart defects among the additional findings of the classic form, and Saudubray associates WFS1 with cardiomyopathy [1, 41]. Respiratory problems result from central apnoea due to brainstem atrophy, bulbar dysfunction with aspiration and sleep disturbance [1]. Chronic fatigue and headaches are frequent, aggravated by infections [3].

3.9. Dominant WFS1 forms

The non-classic forms are less frequent: in a British multidisciplinary service, about 15 of more than 100 patients with WFS1 variants had a dominant form, and at least 14 families had been published by 2022 [1]. The systematic review of 86 patients with Wolfram-like syndrome found variable combinations of optic atrophy, diabetes, deafness, AVP deficiency and endocrine, neurological, urological and psychiatric manifestations; the most frequent phenotype was optic atrophy with deafness (47%), diabetes occurred in 44% and cataracts were frequent; deafness was generally the first manifestation (first decade), followed by diabetes a few years later and optic atrophy in the second decade; the limited follow-up data suggest a non-progressive course without neurodegeneration, although neuroradiological studies are lacking [2, 65]. The families described illustrate the heterogeneity:

In the Washington University registry, visual acuity and retinal nerve fibre layer thickness did not vary significantly with age, which the authors attribute in part to an OCT floor effect [18]. The audiological distinction is useful: in the recessive form deafness begins in the second decade at high frequencies, and in the dominant forms it begins early, at low and middle frequencies [18]. OMIM describes Wolfram-like syndrome by the triad of progressive congenital deafness, diabetes and optic atrophy, with deafness generally diagnosed in the first decade and mainly affecting the low and middle frequencies [25]. In the families of Rendtorff and colleagues, 7 of 8 had spouses with isolated sensorineural deafness, and 3 probands also had a pathogenic GJB2 variant inherited from a deaf parent without optic atrophy; in a British family with p.(Gly780Ser), a 14-year-old girl had profound prelingual deafness and optic atrophy, with normal blood glucose and urine osmolality, her 9-year-old brother also had autism spectrum disorder and the affected mother had schizophrenia [5, 25]. In one of the families with p.(Ala684Val), the proband's asymptomatic wife was heterozygous for p.(Val415del), and the two children, compound heterozygotes, had Wolfram syndrome with juvenile diabetes, optic atrophy, profound deafness from birth or early childhood and congenital cataracts; in HEK cells, the two variants markedly reduced protein expression [5].

3.10. Neonatal or infancy-onset diabetes with deafness and cataracts (de novo dominant form)

De Franco and colleagues identified this syndrome by exome sequencing of a trio and then by screening for heterozygous WFS1 missense variants in 311 patients with diabetes diagnosed before one year of age and without a known cause; they found 5 patients with 3 variants, 4 confirmed de novo [17]. Diabetes appeared before 12 months in all (before 6 months in 2), sensorineural deafness shortly after birth in all, congenital cataracts in 4 and hypotonia in 4 [17].

Source: De Franco and colleagues, 2017 [17]. There are no long-term outcome data; severity is greater than in the classic form, which distinguishes this syndrome from the other dominant forms [1, 2].

3.11. Wolfram syndrome type 2

WS2 was described in families of Palestinian Arab origin with juvenile diabetes, optic atrophy, high-frequency sensorineural hearing loss, urinary tract dilatation, impaired renal function, hypogonadism and severe gastrointestinal ulceration and bleeding; in one family there was facial dysmorphism [1, 32]. The absence of AVP deficiency was taught as a distinguishing feature, but it was later described in patients from Italy and Morocco, as were psychiatric abnormalities and variable optic atrophy; in a subgroup, peptic ulcers, mucocutaneous bleeding and a platelet aggregation defect are observed [1, 38, 51, 66]. ISPAD retains the classic description, without diabetes insipidus and with a bleeding diathesis and peptic ulcer [57]. The largest cohort, with 22 patients from 11 unrelated Palestinian families, all homozygous for the founder variant c.109G>C, broadened the phenotype with frequent psychiatric morbidity and congenital heart disease, in addition to atypical juvenile diabetes and gastrointestinal bleeding [44]. Reduced life expectancy has been described, and the clinical spectrum is not established owing to the small number of patients [1, 38].

In the original description by El-Shanti and colleagues (16 patients from 4 consanguineous Jordanian families), diabetes appeared in the first or second decade, followed by visual loss in all and high-frequency deafness in all except two; eight had urinary tract dilatation, several had severe upper gastrointestinal ulceration and bleeding, and none had diabetes insipidus [6]. In the follow-up of three of these families, one patient with depression died by suicide and others developed symptomatic deafness [6]. The Italian adolescent girl described by Mozzillo and colleagues had recurrent bleeding ulcers at 5 years, optic neuropathy without atrophy at 16, diabetes at 17 and high-frequency deafness, with deficient platelet aggregation in response to ADP; the authors considered that the Jordanian patients might also have had optic neuropathy rather than atrophy [6, 66]. In two Italian sisters described by Rondinelli and colleagues, duodenal ulcers and anaemia began in the first months of life, mild high-frequency deafness was detected at 10 and 13 years, non-autoimmune diabetes requiring insulin appeared at 12 and mild optic atrophy a few years later; the elder had neurogenic bladder with self-catheterisation and the younger mild polyuria since 9 years, without progression, and both had hypogonadism or menstrual abnormalities, hypogammaglobulinaemia and rheumatological disease (HLA-B27-positive spondyloarthritis in one and Löfgren syndrome, a form of sarcoidosis, in the other) [6].

3.12. Natural history and treated disease

The frequencies and ages above are largely from patients treated symptomatically, but without disease-modifying therapy; the available trials have not altered the course [39, 67]. The improvement in survival observed in recent series reflects treatment of complications (hypoglycaemia, urinary infections, central apnoea), not modification of the neurodegeneration [1].

4. Clinical Recognition and Warning Signs

4.1. When to suspect, by age and context

In the infant, diabetes diagnosed before 12 months, especially if accompanied by congenital deafness, congenital cataracts or glaucoma and hypotonia, should raise the possibility of the de novo dominant WFS1 form; ISPAD recommends immediate molecular genetic testing in all infants with diabetes in the first 6 months (grade B) and considering it between 6 and 12 months, especially in the absence of autoantibodies (grade C) [17, 57]. In the child, the central warning sign is the combination of insulin-dependent diabetes without autoantibodies and visual loss due to optic atrophy before 16 years; the absence of autoantibodies in a child labelled as having type 1 diabetes should, on its own, prompt genetic testing, ideally with a broad panel [1, 3]. Visual loss in a child with recent-onset diabetes is not diabetic retinopathy: the duration is too short, and fundoscopy shows optic disc pallor without retinopathy [57, 68]. In the adolescent and young adult, polyuria that persists despite good glycaemic control, new-onset enuresis or incontinence, recurrent urinary infections, delayed puberty in boys and high-frequency hearing loss complete the picture [1, 37]. In the adult, the combination of late-onset diabetes and optic atrophy, with slow progression, may correspond to the adult-onset subtype, frequently with missense variants [3].

The findings suggesting a dominant form are optic atrophy with early-onset sensorineural hearing loss, low-frequency hearing loss in a child, congenital nuclear cataract, diabetes with onset after 16 years and a family history with affected members of both sexes in several generations [1]. The absence of a family history excludes neither form [1]. The GeneReviews criteria for suspicion are summarised in the following table.

Source: GeneReviews, suggestive findings, based on Barrett 1995 and Urano 2016 [1, 69].

4.2. Warning signs and urgent situations

Hyperglycaemia in a child requires same-day referral to a paediatric diabetes specialist, without delaying treatment, to prevent ketoacidosis [37]. Severe hypoglycaemia, more frequent than in type 1 diabetes, can be fatal and is one of the causes of death described [1, 16]. Acute urinary retention in a megabladder, fever with urinary infection due to stasis and obstructive acute kidney injury are urological emergencies [3, 63]. Central apnoea or respiratory failure due to brainstem atrophy, dysphagia with choking and aspiration, status epilepticus and hyponatraemia on desmopressin are other risk situations [1, 37]. Suicidal ideation requires immediate psychiatric assessment, because suicide is among the causes of death [1, 16]. In the patient with corticotrophin deficiency, intercurrent illness and surgery require glucocorticoid cover [2]. In WS2, gastrointestinal bleeding and mucocutaneous bleeding are the specific acute complications [1, 38].

4.3. Newborn screening

The sources consulted do not describe biochemical or genetic newborn screening for Wolfram syndrome. In the dominant forms, deafness is congenital or diagnosed in the first two years (median 2.0 years in the Washington University registry), and early low-frequency sensorineural hearing loss, or hearing loss accompanied by cataracts, should lead to WFS1 being included in the genetic panel [1, 17, 18]. The Oxford Desk Reference recommends, in the investigation of childhood deafness, dipstick testing for glycosuria (for DIDMOAD and for m.3243A>G) and fundoscopy in all, with a search for optic atrophy when diabetes is present [61].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Clinical criteria and molecular definition

The 2014 criteria of EURO-WABB (a European consortium funded by the European Commission) require two major criteria, or one major and two minor criteria, or two pathogenic variants in WFS1 or CISD2; the percentages are the frequencies observed in the EURO-WABB registry of 121 participants with genetic confirmation [37].

Source: EURO-WABB, Table 1 [37]. For Wolfram-like syndrome, EURO-WABB requires one criterion among diabetes (or glucose intolerance), optic atrophy or deafness and at least one loss-of-function variant in WFS1 or CISD2 [37].

The 2019 French PNDS uses the same rule (two major criteria, or one major and two minor) and the same list of criteria, with two differences: neurological signs are entered as “neurological manifestations”, without enumeration, and the presence of a loss-of-function variant does not appear among the minor criteria, because confirmation always requires two biallelic pathogenic variants in WFS1 or CISD2 [23, 37]. Among the suggestive data that do not count as criteria, it adds to those of EURO-WABB nystagmus, hypothyroidism, pubertal delay and, in WS2, gastric disorders, gastrointestinal bleeding and platelet dysfunction [23]. Incomplete recessive forms, with only one major criterion (mainly optic neuropathy without diabetes) and one or more minor criteria, require the same initial assessment and the same follow-up as the complete form, because new manifestations, sometimes subclinical, may appear; the dominant forms undergo the same initial assessment, with an adapted follow-up that the PNDS does not detail [23].

The most recent guidelines have shifted the diagnosis to genetics. GeneReviews establishes the classic form by the presence of suggestive findings and biallelic pathogenic or likely pathogenic variants in WFS1, and the non-classic form by suggestive findings and a heterozygous variant [1]. The Italian consensus defines the syndrome by the presence of biallelic variants classified according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP), in the homozygous (frequent in consanguinity) or compound heterozygous state, with early diabetes and optic atrophy [2]. The international consensus requires biallelic WFS1 or CISD2 variants and accepts, in the childhood-onset subtype, one pathogenic variant with one VUS when the picture is compatible and the decision is taken with a geneticist [3]. No non-molecular test is confirmatory, and the clinical tetrad alone does not establish the diagnosis; it is the number of alleles, and not the phenotype, that defines the form, the counselling and the intensity of surveillance [39]. Before these guidelines, López de Heredia and colleagues, noting that 15% of published patients with two variants did not meet the minimum criteria, proposed broadening the clinical definition to diabetes with optic atrophy at any age, which would cover 96% of patients, or to any two manifestations of the DIDMOAD tetrad, which would cover almost 99%, and preferred the second option [24].

5.2. Biochemical and functional assessment

Diabetes is diagnosed by the ISPAD criteria: random blood glucose of 11.1 mmol/L (200 mg/dL) or more with classic symptoms or a hyperglycaemic crisis; fasting blood glucose of 7.0 mmol/L (126 mg/dL) or more; 2-hour blood glucose in the oral glucose tolerance test (OGTT, 75 g) of 11.1 mmol/L or more; or HbA1c of 48 mmol/mol (6.5%) or more; in the absence of unequivocal hyperglycaemia, two abnormal tests are needed [3, 37]. Autoantibodies (anti-GAD, anti-IA-2, anti-insulin and, if available, anti-ZnT8) are generally negative, and basal C-peptide or C-peptide after a standardised meal (mixed-meal test) documents insulin reserve [2, 37]. In northern Europe, 80% of people with type 1 diabetes have autoantibodies at diagnosis, less frequently more than 5 years later, which limits the value of a late negative result [64].

AVP deficiency is assessed by osmolality and sodium in a morning urine sample paired with fasting plasma, after overnight and morning euglycaemia; the international consensus places the water deprivation test as second line, only in a specialist centre and under supervision, and the Italian consensus adds copeptin where available [2, 3, 37]. Gonadal function is assessed with testosterone or oestradiol, luteinising hormone (LH), follicle-stimulating hormone (FSH), inhibin B and anti-Müllerian hormone (AMH); the thyroid with TSH and free T4; growth with charts and, if indicated, a growth hormone stimulation test [2, 37].

Ophthalmological assessment includes acuity with refraction, colour vision (Ishihara), visual fields, fundoscopy, slit-lamp biomicroscopy, OCT of the nerve fibre layer and macular ganglion cells, OCT-A, visual evoked potentials and pattern electroretinogram [2, 3, 37]. Audiological assessment includes pure-tone and speech audiometry, brainstem auditory evoked potentials and otoacoustic emissions [1]. Brain magnetic resonance imaging, without contrast and with thin sections of the brainstem and orbits, is performed at diagnosis [3]. Urological assessment includes a questionnaire and voiding diary, renal function, renal and bladder ultrasound with post-void residual and urodynamics [3, 37]. When WS2 is suspected, upper gastrointestinal endoscopy and platelet aggregation studies are added [38].

The PNDS details the initial assessment by system [23]. In diabetes, it confirms negativity of anti-insulin, anti-GAD, anti-IA-2 and anti-ZnT8 antibodies and tests for the HLA DR3 and DR4 groups predisposing to type 1 diabetes, whose association is rare [23]. In ophthalmology, it requests refraction under cycloplegia, acuity on the ETDRS (or Monoyer) scale, tonometry, fundoscopy with disc photography, colour vision with the 15-hue test, saturated or desaturated, or with the Lanthony 28-hue test depending on acuity, manual Goldmann visual fields to look for a central or centrocaecal scotoma, autofluorescence to exclude optic disc drusen and retinal dystrophy, peripapillary, macular ganglion cell and macular OCT, electroretinogram and visual evoked potentials, generally normal with flash and with reduced amplitude and increased latency with pattern stimulation [23]. In neurology, it requests clinical examination, brain magnetic resonance imaging, age-appropriate neuropsychological assessment and, if indicated, electromyography and EEG [23]. In urology, validated questionnaires (USP, Urinary Symptom Profile, and ICIQ-FLUTS), urea and creatinine with estimated glomerular filtration rate, renal and bladder ultrasound with post-void residual and urodynamics [23]. In otorhinolaryngology, otoscopy with tympanometry, pure-tone and speech audiometry in quiet and, if possible, in noise, auditory evoked potentials with thresholds, otoacoustic emissions, auditory steady-state responses (ASSR) and cochleovestibular magnetic resonance imaging, which may be performed during follow-up, together with the neurological scan, to avoid general anaesthesia; vestibular assessment (videonystagmoscopy, video head impulse test, videonystagmography with rotatory and caloric tests, vestibular evoked myogenic potentials and a global balance test such as the MiniBEST), smell and taste testing, and swallowing assessment by nasendoscopy, videofluoroscopy and speech and language therapy [23]. In endocrinology, TSH and thyroid hormones, FSH, LH, testosterone and inhibin B [23]. In AVP deficiency, suspicion rests on polyuria above 3 L/day, with serum sodium (or plasma osmolality) at the upper limit of normal and urine osmolality below plasma osmolality; glycosuria, which causes osmotic polyuria with urine osmolality close to plasma osmolality, is the main confounding factor, and the distinction between central and nephrogenic origin is made by the response to desmopressin (dDAVP) in a day hospital, in a centre experienced in polyuria in case of doubt [23]. In WS2, the PNDS recognises the risk of gastrointestinal bleeding from ulcers, aggravated by the platelet defect, but does not recommend systematic coagulation or gastrointestinal assessment [23].

5.3. Genetic tests in order and their limitations

The first-line test is a multigene panel (monogenic diabetes, deafness or retinal and optic nerve dystrophies) that includes WFS1, CISD2 and the genes of the differential diagnosis, with deletion and duplication analysis; exome or genome are alternatives when the phenotype is non-specific [1, 2]. The Italian consensus recommends that all patients with clinical suspicion undergo WFS1 and CISD2 testing by next-generation sequencing, in specialist centres, with results communicated in a genetic counselling consultation and extension to parents and first-degree relatives [2]. EURO-WABB, in 2014, proposed testing CISD2 only after negative sequencing and MLPA (multiplex ligation-dependent probe amplification) of WFS1, in the presence of a characteristic WS2 phenotype or Middle Eastern origin; later guidelines test both genes together [2, 3, 37].

The French PNDS proposes testing WFS1 first and, if negative, CISD2, while accepting simultaneous testing in a panel with both genes; it also recommends systematic testing for WFS1 variants in any isolated optic neuropathy, because visual loss may precede diabetes or appear only in adolescence [23]. In France, the test is performed only after a medical genetics consultation and with the informed consent of the patient or parents; in adults lacking capacity (severe cognitive impairment or psychiatric illness, guardianship), consent requires the signature of the legal representative [23]. According to the PNDS, compound heterozygosity is more frequent than homozygosity, and about 10% of patients with Wolfram syndrome have no identified variant [23].

Sources: GeneReviews, Table 1, and the Italian consensus [1, 2, 45]. The figure of more than 95% is the proportion of known variants that the method detects, not clinical sensitivity. In patients with a clinical diagnosis, both variants are found in 90% [2, 70]; in the series described in OMIM, the rate was 18 of 19 Italian patients, 11 of 14 chromosomes in Danish families and 53 of 59 French patients, in whom 3 had only one variant and 3 had no variant in WFS1 or CISD2 [5, 24]; when only one is found, the explanations are an undetected second variant (promoter, untranslated regions, technical limitations), a dominant variant or another gene; a small proportion of patients with a clinical diagnosis have no identified variants [2, 53]. Apparent homozygosity for a variant without consanguinity may correspond to a deletion on the other allele (false homozygosity) or to uniparental isodisomy, and requires testing of the parents [1]. An example is the Australian family in which three siblings appeared homozygous for p.(Pro504Leu), inherited from the father, while the mother did not carry the variant, which led to the assumption of a deletion on the maternal chromosome [5]. A negative result does not exclude the diagnosis, and parental mosaicism may not be detected in leucocyte DNA [1].

5.4. Report, VUS and segregation

The report should state the gene, the versioned transcript (NM_006005.3 or NM_001008388.5), HGVS (Human Genome Variation Society) nomenclature at DNA and protein level, zygosity, the ACMG/AMP class and, when there are two variants, whether they are in trans [2, 8]. In the clinical setting, the pathogenic and likely pathogenic classes are equivalent; a VUS neither confirms nor excludes the diagnosis and should not underpin clinical decisions; VUS should be reviewed periodically, for example once a year, at the clinician's request to the laboratory [1, 2]. A heterozygous WFS1 finding does not predict AVP deficiency, neurodegeneration, neurogenic bladder or reduced survival, and should not be reported as Wolfram syndrome [39]. Hôpital Necker registered in 2026 a study on VUS and on the distinction between the dominant Wolfram-like form and the recessive form in patients with macular abnormalities or cysts on OCT (NCT07485413) [55].

5.5. Practical algorithm

In the presence of diabetes without autoantibodies with optic atrophy, or neonatal diabetes with deafness and cataracts, or optic atrophy with early deafness, the first step is the multigene panel with copy number analysis and, in parallel, the baseline assessment by system described above [1, 2]. If two variants are found, phase is confirmed by testing the parents and multisystem surveillance of the classic form is started [1, 37]. If one variant is found, it is assessed whether it is a described dominant variant, whether it is de novo or shows compatible family segregation, and the second variant is sought by copy number analysis and, if necessary, by genome sequencing [2]. If the panel is negative and the phenotype convincing, deletion and duplication analysis and exome or genome reanalysis are performed, and the differential diagnoses are considered, including m.3243A>G and OPA1 [39, 64]. Fig. 2 of the Italian consensus presents a diagnostic roadmap; the image was not consulted [2].

5.6. Prenatal and preimplantation diagnosis

Once the family variants have been identified, prenatal diagnosis and preimplantation genetic testing are possible for both the classic form and the dominant forms; the first molecular prenatal diagnosis of the syndrome was reported in 2004 [1, 37]. No specific fetal ultrasound findings are described in the sources consulted.

6. Differential Diagnosis

6.1. Comparative table

Sources: GeneReviews (Tables 3 and 4), Italian consensus (Table 4), EURO-WABB, GeneReviews on permanent neonatal diabetes and on deafness, Oxford Desk Reference and Saudubray [1, 2, 37, 41, 60, 61, 64, 73–75].

6.2. Diabetes with deafness and ocular involvement

The main differential diagnosis of the classic form is mitochondrial disease. Saudubray includes Wolfram syndrome in the table of recognised mitochondrial syndromes, alongside MIDD, in which sensorineural deafness and insulin-dependent diabetes generally begin in the fourth decade, with macular dystrophy, cardiomyopathy with a risk of sudden death and nephropathy [74]. The Oxford Desk Reference recommends testing for m.3243A>G when there is matrilineal transmission or deafness, and notes that in mitochondrial diabetes deafness usually precedes diabetes, unlike in Wolfram syndrome [64]. Thiamine-responsive megaloblastic anaemia shares non-autoimmune diabetes, early deafness and, in case reports, optic atrophy; the full blood count and the glycaemic response to thiamine in childhood guide the diagnosis [1, 2]. The ciliopathies (Alström and Bardet-Biedl) are distinguished by retinal rather than optic nerve dystrophy, by obesity and by diabetes with insulin resistance [1]. Friedreich ataxia may combine diabetes, optic atrophy and deafness, but ataxia and areflexia dominate the picture and the GAA expansion confirms the diagnosis [1]. The table of associations between syndromes and diabetes in the Oxford Desk Reference places Wolfram syndrome in the columns for deafness, ocular signs and kidney disease [64].

The overlap with TRMA explains a historical error: two Italian children described in 1989 as having DIDMOAD with megaloblastic and sideroblastic anaemia, neutropenia and a haematological and glycaemic response to thiamine turned out, in 1997, to be cases of TRMA linked to chromosome 1q, whereas patients with the classic tetrad without anaemia did not respond to thiamine [24]. OMIM also records a possible mitochondrial form of Wolfram syndrome (598500): a girl with insulin-dependent diabetes from early childhood, progressive deafness and optic atrophy and mild hyperlactataemia had a heteroplasmic 7.6 kb deletion of mitochondrial DNA, absent in the parents, and two Spanish families with linkage to 4p16 had multiple mitochondrial DNA deletions, up to 85 to 90% in the central nervous system of one patient [24]. The relationship between these deletions and WFS1 variants has not been clarified, and the British series of 19 families and the Lebanese series found no pathogenic mitochondrial DNA alterations [5, 24]. Langerhans cell histiocytosis is an acquired cause of diabetes mellitus associated with diabetes insipidus [24].

6.3. Optic atrophy with deafness without diabetes

When there is no diabetes, the differential diagnosis of dominant Wolfram-like syndrome is mainly OPA1-related dominant optic atrophy, which is accompanied by auditory neuropathy in some patients [1, 2]. The dominant syndromic form of OPA1 (OMIM 125250) combines optic atrophy with deafness, ophthalmoplegia, myopathy, ataxia and neuropathy [72]; in the families with Wolfram-like syndrome described in OMIM, testing for OPA1, TIMM8A and the most frequent Leber neuropathy variants had been negative before the WFS1 variant was identified [25]. The Oxford Desk Reference recommends looking for optic atrophy if there is diabetes (WFS1) or auditory neuropathy (OPA1) [61]. The X-linked forms (PRPS1, TIMM8A) mainly affect males and are associated with peripheral neuropathy or dystonia [1]. The audiometric pattern helps: low-frequency hearing loss points to WFS1 or DIAPH1, and lamination of the outer plexiform layer on OCT is a characteristic finding of the dominant WFS1 forms [18, 61]. Leber hereditary optic neuropathy, maternally inherited, is another differential diagnosis listed by EURO-WABB [37]. The chronology of visual loss also distinguishes them: Leber neuropathy begins abruptly, progresses rapidly and predominates in young adult men, whereas dominant optic atrophy and Wolfram syndrome begin insidiously in childhood, without sex predominance; according to the discussion by Cunha and colleagues, which summarises the literature, the RNFL decreases faster in Wolfram syndrome than in dominant optic atrophy, and the macular GCC is thin from early on in dominant optic atrophy and only later in Wolfram syndrome [52]. Dominant optic atrophy belongs to the group of hereditary optic atrophies.

The PNDS places in the differential diagnosis of the dominant syndromic WFS1 forms with progressive deafness and optic neuropathy dominant optic atrophy with deafness due to a heterozygous OPA1 variant (OPA1 plus form) and the OPA3 gene, and notes that neurological manifestations are rarer in these dominant forms than in the classic form [23].

6.4. Neonatal diabetes and adult-onset diabetes

In neonatal diabetes, GeneReviews lists classic (recessive) Wolfram syndrome among the syndromic differential diagnoses, and the de novo dominant WFS1 form adds congenital deafness and cataracts [17, 73]. Saudubray includes IER3IP1, PTF1A and WFS1 among the causes of neonatal hyperglycaemia associated with microcephaly and brain malformations, and Wolcott-Rallison syndrome (EIF2AK3) among the diabetes forms due to endoplasmic reticulum stress [41, 75]. Most neonatal diabetes results from de novo heterozygous variants in non-consanguineous families, and the recessive forms predominate in consanguineous families [64]. In the adult, dominant WFS1-associated diabetes is rare and should be distinguished from MODY (maturity-onset diabetes of the young; 1 to 2% of people with diabetes), in which GCK causes stable hyperglycaemia of 5.5 to 8 mmol/L and HNF1A and HNF4A respond to low-dose sulphonylureas [3, 35, 64]. These differential diagnoses include monogenic neonatal diabetes, MODY-type monogenic diabetes and HNF1B-related disease, including the 17q12 deletion.

6.5. Distinguishing WS1 from WS2

The clinical distinction between WS1 and WS2 is imperfect, and GeneGuidance considers WS2 clinically inseparable from the classic WFS1 form [38, 39]. Peptic ulcer, bleeding and the platelet aggregation defect are the clues to WS2; Palestinian or Jordanian origin and consanguinity increase the probability, but there are patients in Italy, Morocco and France [1, 44, 51]. Confirmation is molecular [2].

7. Treatments and Risk Management Strategies

7.1. Principles and organisation of care

There is no cure and no therapy that has been shown to alter the course of the disease; treatment is symptomatic and organised by a multidisciplinary team [1, 39]. The international consensus defines a core team with a coordinator (geneticist or nurse navigator), endocrinologist, neurologist, otorhinolaryngologist or audiologist, ophthalmologist or neuro-ophthalmologist with optometry and urologist with access to urodynamics, and an extended team with a genetic counsellor, psychologist or psychiatrist, speech and language therapist, physiotherapist, gastroenterologist, transition coordinator, sleep specialist and a link to a patient association [3]. The Italian consensus adds nephrology, internal medicine as coordinator in the adult, intensive care for acute complications, neuroradiology and liaison with general practitioners [2]. The international consensus recommends early integration of palliative care and advance care planning into the multidisciplinary review [3]. The recommendations of the three specific guidelines (EURO-WABB 2014, SID/SIEDP 2025 and the 2026 international consensus) are based on observational studies and expert opinion; no recommendation has grade A, because there are no randomised trials with published results [2, 3, 37].

The 2019 French PNDS, prepared by the OPHTARA, MAOLYA and CALISSON reference centres with the Haute Autorité de Santé and taking the European recommendations into account, entrusts coordination to a physician from the reference or competence centre, in liaison with the attending physician [23]. It provides for a dedicated consultation to disclose the diagnosis, to explain the disease, the natural history, the prognosis, the care plan and social and medico-social support, to introduce the patient association and to offer genetic counselling; in children and adolescents, preferably with both parents and, if possible, with a psychologist, social worker and specialists, followed by a second consultation to answer the questions that arise afterwards [23]. It sets as objectives early control of diabetes and prevention of its complications, treatment of endocrine, urological and neurological complications, compensation for sensory and balance deficits and school and occupational inclusion; the list of professionals also includes a gastroenterologist, pulmonologist, electrophysiologist, physiatrist, orthoptist, psychomotor therapist, occupational therapist, dietitian and social worker, and liaison with establishments for people with visual, hearing or motor impairment [23]. The aims of follow-up are to compare each system with the previous assessment, detect involvement of new organs and diabetic and neurological complications, assess the efficacy, tolerability and adherence to treatments and measure the school, occupational and psychological impact [23].

7.2. Diabetes mellitus

Insulin is the mainstay of treatment, following current practice for insulin-dependent diabetes [1, 3]. Because severe hypoglycaemia is frequent, GeneReviews recommends multiple daily injections or continuous subcutaneous insulin infusion by pump, with continuous glucose monitoring (CGM) with a hypoglycaemia alarm [1, 58]. The international consensus considers CGM integrated into a hybrid closed-loop system the preferred method when feasible, with audible and vibrating alarms adapted to visual loss; the Italian consensus places advanced hybrid closed-loop systems and multiple daily injections as equivalent options, to be chosen according to lifestyle [2, 3]. The insulin dose is individual; in the cohort of Cano and colleagues, the mean was 0.71 U/kg/day [42]. EURO-WABB recommends an HbA1c target below 7.5%, less strict in the presence of severe hypoglycaemia, hypoglycaemia unawareness, limited life expectancy or extensive comorbidities, quarterly HbA1c, self-monitoring devices adapted to low vision and nutritional assessment at least annually [37]. With a chronically elevated HbA1c above 7.5%, especially in adolescence, psychological distress, eating disorders and insulin manipulation should be sought [37]. Ketoacidosis is rare at presentation but may occur in advanced stages, and it is treated by standard paediatric protocols, with attention to the risk of cerebral oedema [37].

Severe hypoglycaemia is treated in hospital with intravenous dextrose and at home with subcutaneous glucagon, followed by oral glucose [37]. The guideline does not give the dose; the US label for injectable glucagon (Glucagon®) indicates 1 mg subcutaneously, intramuscularly or intravenously in adults and children weighing 20 kg or more, and 0.5 mg (or 20 to 30 µg/kg) below 20 kg, with a second identical dose if there is no response within 15 minutes [76]. In Wolfram syndrome, hypoglycaemia awareness may be severely impaired, which justifies less strict targets and CGM [37].

GeneReviews states that other antidiabetic drugs are not authorised for Wolfram syndrome owing to insufficient evidence of efficacy and safety [1]. Glucagon-like peptide-1 (GLP-1) receptor agonists have been studied for their effects on the β cell and possible neuroprotection. In an Italian series of 4 children aged 10 to 14 years, treated off-label for 8 to 27 months, liraglutide (Victoza®) was given subcutaneously with the evening meal, as an adjunct to insulin, at a starting dose of 0.6 mg/day (0.3 mg/day in two children) increased up to 1.8 mg/day; all reached and maintained the maximum dose, with mild and transient gastrointestinal symptoms, a C-peptide area under the curve between 81% and 171% of baseline and stable neuro-ophthalmological parameters [77]. The US label for Victoza® indicates, in type 2 diabetes in adults and in children from 10 years, 0.6 mg/day for at least one week, then 1.2 mg and 1.8 mg/day, and the EMA has authorised it since 30 June 2009 in the same population; the regimen of the series matches that of the label, but the indication does not [78, 79]. In a retrospective cohort from the Washington University registry, 30 of 84 patients received a GLP-1 agonist (semaglutide, liraglutide, dulaglutide or tirzepatide); HbA1c did not change at 1 year (−0.02%; 95% CI −0.56 to 0.51; n=12) or at 2 years (+0.14%; 95% CI −0.31 to 0.59; n=14), visual acuity worsened in a manner consistent with natural progression (+0.12 logMAR at 2 years; 95% CI 0.006 to 0.24), there were adverse effects in 56.7% (gastrointestinal in 50%) and 13 patients discontinued [80]. In a New Zealand series, 5 treated patients had a mean reduction in HbA1c of 11.6 mmol/mol and in insulin of 0.25 U/kg/day, with short follow-up [81]. The two studies are observational and diverge. The international consensus considers GLP-1 agonists an adjunctive option, which does not replace insulin and should not replace it completely in children, and calls for the gastrointestinal effects to be weighed against dysautonomia and growth delay [3]. Liraglutide has orphan designation in the EU for Wolfram syndrome (EU/3/22/2667) [82].

The French PNDS treats diabetes with insulin in a basal–bolus regimen or by pump, notes that the glycaemic target is easier to achieve at first, thanks to residual secretion, and that with age blood glucose becomes less predictable, while neurological and cognitive deficits make the use of pumps and sensors more difficult and increase dependence on relatives [23]. It considers GLP-1 receptor agonists promising, because of their effect on reticulum stress and β-cell survival, and proposes starting them early in the course; it cites a case of WS1 in which exenatide reduced insulin requirements by 70% and increased basal insulin secretion sevenfold, and DPP-4 inhibitors as an oral alternative, of lower expected efficacy [23]. Secretagogues have shown no consistent benefit in murine models, and gliclazide was ineffective in one study; pioglitazone had a favourable effect on blood glucose and β-cell survival in a 2009 study, but was withdrawn from the French market in 2011 because of the risk of bladder cancer and osteoporosis [23]. At the time of the PNDS, GLP-1 analogues had no paediatric authorisation in France; liraglutide is now authorised in the EU for type 2 diabetes from 10 years, which does not change its off-label status in Wolfram syndrome [23, 79]. The PNDS proposal for early initiation predates the negative observational data from the North American cohort and rests on a single case [23, 80].

The complications of diabetes are monitored as in type 1 diabetes: EURO-WABB recommends annual screening for nephropathy from 12 years with diabetes for more than 5 years (albumin/creatinine ratio) and starting an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker as soon as microalbuminuria is confirmed; annual retinopathy screening with retinal photography after 5 years of diabetes; annual neurological examination for neuropathy; lipid profile at 12 and 17 years (or earlier, with risk factors); and blood pressure measurement at least annually [37]. The sources consulted do not indicate a specific drug or dose for these complications in Wolfram syndrome.

7.3. Arginine vasopressin deficiency

First-line treatment is desmopressin by the oral, sublingual or intranasal route, with an individualised dose and monitoring of serum sodium, and attention to fluid balance during intercurrent illness [3]. EURO-WABB asks that bladder dysfunction always be considered before increasing the dose, because desmopressin carries a risk of hyponatraemia, and GeneReviews warns of the risk of hyper- and hyponatraemia [1, 37]. EURO-WABB, the Italian consensus and the international consensus do not give a numerical dose; the only specific guideline with a dose is the French PNDS, and the label serves for comparison. In Portugal, the SmPC for oral lyophilisate Minirin® (60 and 120 µg) indicates in AVP deficiency a sublingual starting dose of 60 µg three times daily in adults and children, maintenance of 60 to 120 µg three times daily and a usual total daily dose of 120 to 720 µg; it is stopped if there are signs of fluid retention or hyponatraemia; it is contraindicated in habitual or psychogenic polydipsia (more than 40 mL/kg/24 h), in heart failure or on diuretics, with creatinine clearance below 50 mL/min, in hyponatraemia, in the syndrome of inappropriate antidiuretic hormone secretion and in von Willebrand disease type IIB [83]. In the USA, DDAVP® tablets start at 0.05 mg twice daily, with an optimal dose of 0.1 to 0.8 mg/day in divided doses (range 0.1 to 1.2 mg/day), used safely from 4 years; the nasal spray uses 10 to 40 µg/day in the adult and 10 µg once daily up to 30 µg/day in the child [84, 85]. The doses differ between the tablet, the sublingual lyophilisate and the nasal spray, and are not directly convertible between formulations [83, 84]. The international consensus notes that desmopressin can also reduce urine output and improve continence in neurogenic bladder [3]. In Brazil, DDAVP®, Eponim®, Dosyx®, FURP-Desmopressina and generic desmopressin acetate are registered [86].

The French PNDS treats AVP deficiency only when the functional impact of polyuria or frequency, possibly nocturnal, is excessive; if it is tolerable, surveillance suffices [23]. It recommends starting desmopressin preferably as an inpatient, with weight and serum sodium (or plasma osmolality) measured several times a day, and lists the French formulations: Minirin® nasal spray with 10 µg per spray and Minirinmelt® orodispersible tablets of 60, 120 and 240 µg [23]. At first requirements are generally low and the response rapid, and two daily doses of 60 µg, morning and evening, may suffice; over time, the dose increases through the tablet strength or the number of doses, up to 3 to 4 per day, and the oral form may be combined with the nasal form; the PNDS sets no maximum dose [23]. This starting dose, by the orodispersible oral route, is lower than that of the Portuguese SmPC for the oral lyophilisate (60 µg three times daily); the guideline dose is the reference and the SmPC dose serves for comparison [23, 83]. As long as thirst and access to water are preserved, the patient regulates intake; because visual and motor impairment may hinder access to water, this should be asked about in the history and facilitated, and at school children and adolescents should have free access to water and toilets, so that they do not restrict their intake [23]. During follow-up, the reappearance of polyuria or new symptoms, such as headaches, require serum sodium to be measured, because changes in AVP deficiency, renal function or diabetes may lead to desmopressin overdosage; because hyperglycaemia also increases thirst, frequent glucose self-monitoring helps to distinguish the two causes [23].

7.4. Other endocrinopathies

Male hypogonadism. EURO-WABB indicates, below 18 years, intramuscular testosterone enanthate in increasing doses of 50 to 250 mg every 3 to 4 weeks; above 18 years, intramuscular testosterone undecanoate every 3 months or testosterone gel 50 mg/day [37]. The guideline gives neither the undecanoate dose nor maximum doses. For comparison, the FDA label for Aveed® (testosterone undecanoate, adults only) indicates 750 mg intramuscularly at initiation, at 4 weeks and then every 10 weeks, without titration, with 30 minutes of observation after each injection because of the risk of pulmonary oil microembolism and anaphylaxis; the FDA label for 1% testosterone gel (NDA021015) starts at 50 mg/day and is adjusted by serum testosterone up to a maximum of 100 mg/day in adult men. The safety and efficacy of both formulations below 18 years are not established in the labels, because of the risk of bone age acceleration and premature epiphyseal closure [87, 88]. The Birmingham and Münster series recommends monitoring Tanner stages, LH, FSH, testosterone or oestradiol and inhibin B from early adolescence, so as to treat early and allow fertility preservation before testicular function is lost [62].

Female hypogonadism. EURO-WABB recommends oestrogen–progestogen replacement without specifying a regimen [37]; the sources consulted do not describe specific doses for the syndrome.

Hypothyroidism. EURO-WABB indicates levothyroxine with a starting dose of 25 µg/day; the original document writes “25 mg/day”, a typographical error in the unit; the guideline gives neither a maintenance dose nor a maximum dose [37]. The fixed dose does not distinguish age or weight: the US label for Synthroid® indicates daily starting doses per kilogram by age (5–6 µg/kg/day from 1 to 5 years, 4–5 µg/kg/day from 6 to 12 years and 2–3 µg/kg/day above 12 years with incomplete growth and puberty, titrating every 2 weeks by TSH and free T4) and, with growth and puberty complete, 1.6 µg/kg/day, titrated in 12.5–25 µg steps every 4–6 weeks, with lower doses and titration every 6–8 weeks in the elderly and in heart disease [89].

The PNDS recommends annual screening for endocrine deficiencies, oral levothyroxine taken in the morning on an empty stomach, with the dose adjusted to TSH, and, in the boy with hypogonadism, testosterone to resume pubertal virilisation, which also improves the well-being of adolescents; it considers that hormone replacement adjusted to each patient may slow some neurological symptoms, a statement without cited comparative data [23].

Growth hormone deficiency. GeneReviews recommends assessing supplementation according to national guidelines [1]. In the absence of a specific dose, the reference is the label: the Portuguese SmPC for Genotropin® (somatropin) indicates, in growth disturbance due to insufficient growth hormone secretion in the child, 0.025 to 0.035 mg/kg/day (0.7 to 1.0 mg/m²/day) subcutaneously, and warns that somatropin may reduce insulin sensitivity and require adjustment of the insulin dose in patients with diabetes [90]. This interaction is relevant in insulin-treated diabetes.

Corticotrophin deficiency. The Italian consensus recommends considering glucocorticoid supplementation in situations of stress and intercurrent illness in patients with anterior pituitary hypofunction, without specifying a regimen [2]. For comparison, the label for Solu-Cortef® (hydrocortisone sodium succinate) in the USA indicates an initial intravenous dose of 100 to 500 mg depending on the situation, to be individualised [91]; stress-dose regimens for secondary adrenal insufficiency are not given in the sources consulted.

7.5. Vision

No proven treatment slows visual loss [3]. Management includes correction of refractive error, low vision consultation with magnifiers, digital and voice systems, registration as severely sight impaired where applicable, orientation and mobility training, Braille, occupational rehabilitation and special educational support [1–3, 37]. Cataract surgery is performed when indicated [37]. The Italian consensus mentions idebenone, proposed to improve mitochondrial function, and states that its efficacy in the optic atrophy of Wolfram syndrome is not established [2]. Idebenone (Raxone®) has been authorised by the EMA since 8 September 2015 only for visual impairment in adolescents and adults with Leber hereditary optic neuropathy, at the recommended dose of 900 mg/day (300 mg three times daily), with no distinct maximum dose defined in the source; use in Wolfram syndrome is off-label and an extrapolation from LHON; idebenone has no FDA authorisation and no registration in the ANVISA open database [86, 92].

The PNDS mentions idebenone, tested in three patients with results that do not allow conclusions about efficacy [23]. It organises visual compensation into refractive correction, filter lenses for photophobia, optical aids (magnifying lenses, magnifiers, telescopes, video magnifiers), non-optical aids (large-print books, large screens, tablets), school support with a classroom assistant, extra time in examinations and electronic devices, and rehabilitation, in a specialist centre or as an outpatient, when acuity and field deteriorate (mobility lessons with a cane, Braille, occupational therapy for spatial orientation, computing and voice software, psychological support), with the aim of preserving social integration in the blindness phase [23]. It recommends assessing visual and auditory function before any rehabilitation, because dual sensory impairment is frequent [23].

7.6. Hearing

There is no treatment that reverses deafness; intervention depends on the degree: hearing aids for mild to moderate loss, cochlear implant for severe loss, speech and language therapy and sign language [1, 2]. The international consensus generally reserves cochlear implantation for the early profound deafness of the dominant forms, and not for childhood- or adult-onset Wolfram syndrome; the Italian consensus proposes it for significant losses, which usually occur in homozygous patients [2, 3]. In the Washington University registry, 8 of 15 patients with a dominant form had bilateral cochlear implants, with a median of 2.2 years between diagnosis and the first hearing aid [18]. The GeneReviews on deafness recommends avoiding repeated noise exposure, with no established safe level; the permissible exposure limit set by OSHA and NIOSH is 85 dB over an 8-hour working day [60].

In the presence of significant early deafness in a child, the PNDS prefers, together with the parents, an auditory–oral or bilingual approach to an exclusively visual–gestural approach, because the expected visual loss and neurological signs will later make the use of sign language difficult [23]. Rehabilitation includes conventional hearing aids, cochlear implant if there is an audiological indication, remote microphones and frequency-modulated systems, interfaces for television and mobile phone and a hearing loop; lip-reading, cued speech and signed French are supports, while maintaining the attempt at autonomous communication through the auditory channel [23]. In the child, speech and language therapy works on auditory training, phonemes, speech and language; in the adult, it focuses on auditory training and memory (mental supplementation), without favouring visual aids, with home sessions if mobility is difficult and a duration adapted to fatigue and concentration [23]. Adults' hearing aids should be checked at the place of residence (home or institution), and vestibular rehabilitation focuses on proprioception, which neurological disease may also impair [23].

7.7. Nervous system, swallowing and breathing

Neurological follow-up includes assessment of ataxia with validated scales (SARA, Scale for the Assessment and Rating of Ataxia), physiotherapy for gait and fall risk, otorhinolaryngological assessment if vestibular involvement is suspected and rehabilitation of disabling nystagmus and cerebellar tremor [3, 37]. Dysarthria and dysphagia are treated with speech and language therapy, modification of food consistency, head positioning and safe swallowing exercises, with oral hygiene; weight loss, respiratory infections and choking lead to referral to speech and language therapy and gastroenterology [1, 3]. Central apnoea is managed with the respiratory team (nocturnal non-invasive ventilation; in the advanced stage, tracheostomy and ventilation); in the presence of respiratory symptoms, EURO-WABB recommends bronchoscopy to assess vocal cord mobility and exclude obstruction, spirometry and morning blood gases [1, 37]. Neuropathic pain is treated with antiepileptic drugs, antidepressants, lidocaine plasters or transcutaneous electrical nerve stimulation (TENS), and epilepsy with antiepileptic drugs after EEG; the sources do not indicate drugs or doses specific to the syndrome [37]. Dysautonomia is managed with conservative measures, and chronic fatigue by treating anxiety, depression, anaemia, sleep and thyroid problems and with supervised graded aerobic exercise [3]. Headaches are treated by correcting dehydration, infection, poor glycaemic control and poor sleep hygiene [3].

The PNDS adds walking aids (walker, orthoses, wheelchair), non-invasive ventilation discussed with pulmonology in sleep apnoea, gastrostomy or enteral nutrition in dysphagia, medical treatment of constipation and, if necessary, a stoma, and, in the presence of associated respiratory problems, discussion of nocturnal ventilatory assistance, possibly with tracheostomy, within the multidisciplinary care plan and with palliative care support; neuropathic pain and epilepsy are treated as in other neurological diseases, and the neurological treatment is not specific to the syndrome [23]. School and occupational adaptations include a classroom assistant, a computer, educational guidance, support for job placement, workplace adaptation and occupational reorientation in the adult [23].

7.8. Neurogenic bladder and kidney

Management involves urology, physical and rehabilitation medicine and neurology [37]. The international consensus proposes a therapeutic ladder: bladder training with double voiding and oral drugs; desmopressin to reduce output; with progression to megabladder and refractory incontinence, clean intermittent self-catheterisation; then an indwelling urinary catheter or a Mitrofanoff appendicovesicostomy [3]. EURO-WABB adds botulinum toxin, electrical stimulation, physiotherapy and surgery according to the specialist, urine culture if there is fever or symptoms, prior assessment of the ability to self-catheterise (ataxia, low vision, cognitive impairment) and the notion that an indwelling catheter is a risk factor for infection [37]. GeneReviews notes that detrusor dyssynergia may require relaxants or stimulants at different stages and asks that AVP deficiency be screened for and treated at the same time [1].

The French PNDS agrees with EURO-WABB on anticholinergics, botulinum toxin, physiotherapy, neurostimulation and intermittent self-catheterisation after assessment of ability in the face of ataxia, cognitive impairment and low vision, and lists the surgical alternatives: temporary diversion by suprapubic catheter or indwelling urethral catheter, permanent diversion and, in bladder acontractility with complications, a Bricker-type cutaneous ileal conduit [23]. It recommends systematic urine culture in the presence of fever or suggestive symptoms and a search for risk factors for infection in catheter carriers [23].

Anticholinergics target the detrusor overactivity phase; in a bladder with obstruction or retention they may precipitate urinary retention, a contraindication stated in the label [93]. The guidelines do not indicate a drug or dose. The Portuguese SmPC for Ditropan® (oxybutynin hydrochloride, 5 mg) indicates it in reflex or neurogenic bladder and, from 5 years, in detrusor overactivity of neurogenic cause; in the adult, 5 mg two or three times daily (maximum 5 mg four times daily); above 5 years, 5 mg twice daily (maximum 15 mg/day); it is contraindicated in bladder outflow obstruction with a risk of retention, intestinal atony and angle-closure glaucoma, and calls for caution in autonomic neuropathy and because of the risk of cognitive impairment and heat prostration due to reduced sweating [93]. In the USA, extended-release oxybutynin uses 5 to 10 mg once daily in the adult (maximum 30 mg/day) and 5 mg once daily from 6 years in detrusor overactivity of neurological cause (maximum 20 mg/day) [94]. Use in Wolfram syndrome is an extrapolation from neurogenic bladder of other neurological causes. The contraindications and precautions in the label coincide with manifestations of the syndrome (described glaucoma, dysautonomia with impaired thermoregulation, cognitive decline), which requires periodic reassessment. In the Italian case described in section 10, the anticholinergic was stopped during preconception planning [42].

7.9. Bowel

Faecal incontinence first requires exclusion of encopresis due to constipation with overflow; this is followed by a diet with fibre and water, a bowel regimen, pelvic floor physiotherapy and gastroenterology if the cause is neurogenic [3]. Constipation is treated stepwise with dietary and behavioural measures, softeners, fibre, laxatives and disimpaction [3]. Bowel management is coordinated with bladder management [1].

7.10. Mental health

All newly diagnosed patients should be assessed by a psychologist or psychiatrist (depression, anxiety, hallucinations or psychosis, attention deficit hyperactivity disorder, obsessive–compulsive symptoms) [3]. Anxiety and depression are treated with psychotherapy and selective serotonin reuptake inhibitors (SSRIs) with a psychiatrist; limited evidence, from case series, suggests that fluoxetine and citalopram may be more effective than sertraline, and the other SSRIs are acceptable [3]. The consensus does not indicate doses. In the US label for Prozac® (fluoxetine), major depression is treated in the adult with 20 mg/day in the morning, up to a maximum of 80 mg/day, and in children and adolescents with 10 or 20 mg/day (10 mg/day as the starting and target dose in lower-weight children) [95]. In the EU, fluoxetine has been authorised since 2006 from 8 years in moderate to severe major depressive episodes that do not respond to 4 to 6 sessions of psychotherapy, always combined with psychotherapy; it starts at 10 mg/day and may be increased to 20 mg/day after 1 to 2 weeks, with monitoring of suicidal behaviour, growth and puberty [96]. The label for Celexa® (citalopram) authorises it only in adults, at 20 mg once daily with a maximum of 40 mg/day because of the risk of QT interval prolongation, and it is not approved in the paediatric age group [97]. The Celexa® label has a boxed warning, common to antidepressants, about the increased risk of suicidal thoughts and behaviour in children and young adults, which adds to the suicide risk inherent in the syndrome and requires close monitoring [1, 97]. SSRIs can cause hyponatraemia through inappropriate antidiuretic hormone secretion, with cases of sodium below 110 mmol/L, which matters in patients on desmopressin [95].

The PNDS proposes systematic psychological support at disclosure of the diagnosis, extended over time, because progression may reactivate the impact of disclosure, and to screen for behavioural, anxiety and mood disorders [23]. It notes the increased risk of discrimination and harassment in private and working life, suggests acceptance and commitment therapy (ACT), whose efficacy on self-determination and self-esteem has been shown in other chronic diseases but not in Wolfram syndrome, possibly with mindfulness-based psychotherapy for the anxiety linked to the unpredictability of symptoms, and recommends family psychological support, because of the guilt, helplessness and exhaustion frequent in family carers [23].

7.11. Non-pharmacological interventions, education and social support

Source: Italian consensus, Table 6, supplemented by GeneReviews [1, 2]. GeneReviews describes, for the USA and the United Kingdom, the individualised education plan reviewed annually, the involvement of vision and hearing consultants at school and school adaptations (front-row seating, assistive technology, scribe, extra time, enlarged text) [1]. The international consensus recommends, at diagnosis, contact with a patient association and assessment of social benefits, carer support and educational support [3].

The French PNDS adds adaptation of food texture in dysphagia and therapeutic education focused on glycaemic control, adjustment of insulin and diet, with a dietitian from the first visit, and on the use of visual and auditory rehabilitation equipment; patient information covers the symptoms that should prompt a consultation, the treatments and their adverse effects and the schedule of follow-up examinations [23].

7.12. Particular features of Wolfram syndrome type 2

In WS2, organ-based treatment is the same, with specific additions: surveillance and treatment of peptic ulcer and gastrointestinal bleeding, and attention to the platelet aggregation defect before surgery and invasive procedures and when considering antiplatelet agents or anticoagulants [38]. No specific protocol or trial has been published [38]. The proof of concept with deferiprone and N-acetylcysteine is described in subsection 7.17.

7.13. Pregnancy, peripartum and lactation

Women with the classic form generally retain the ability to conceive; a 2025 review brought together 20 successful pregnancies published worldwide, all ending by caesarean section between 32 and 39 weeks [1, 42]. The maternal complications described include pre-eclampsia, new or worsened AVP deficiency, hearing loss, acute urinary retention, hydroureteronephrosis, cerebral infarction with hemiplegia after severe AVP deficiency and hypovolaemia, and appendicitis; the fetal complications include polyhydramnios, intrauterine growth restriction and abnormal cardiotocography [42]. Pregnancy in a woman with insulin-dependent diabetes increases the risk of congenital anomaly 2- to 8-fold (craniofacial, cardiovascular, gastrointestinal, urogenital, musculoskeletal and central nervous system diabetic embryopathy); optimising glycaemic control before and during pregnancy reduces this risk without eliminating it, and high-resolution fetal ultrasound, fetal echocardiography and maternal–fetal medicine follow-up are recommended [1]. Because AVP deficiency may appear during pregnancy, it should be monitored [1]. The Minirin® SmPC describes 53 pregnancies exposed to desmopressin in women with AVP deficiency without adverse effects [83]. The following table summarises the ten pregnancies with individual data from the 2025 review.

Source: Caretto and colleagues, 2025, Table 2 [42]. In the 1995 British cohort, five women had nine unaffected children [42].

The peripartum period requires multidisciplinary planning. The checklist published with a 2026 Spanish case includes: an intrapartum and postpartum glycaemic protocol with minimal fasting; adjustment of fluids and desmopressin to fluid balance, with urinary catheterisation and hourly urine output; prevention of urinary infection; assessment of orthostasis and of the autonomic response to stress; titratable neuraxial anaesthetic techniques, preparation for a difficult airway and an alternative plan; prevention of neuraxial hypotension and capnography with oxygen; bipolar electrosurgery if there is a cochlear implant; a communication plan adapted to deafblindness, with the partner present; and thromboprophylaxis and metabolic and fluid monitoring post partum [49]. The same authors recommend minimising neuromuscular blocking agents with train-of-four monitoring, considering avoiding suxamethonium, preventing aspiration in bulbar dysfunction with awake extubation, and note that self-catheterisation may reach about 12 times a day in pregnancy [49]. The sources consulted do not describe data on lactation in Wolfram syndrome.

7.14. Perioperative care and anaesthesia outside pregnancy

The same precautions apply to any anaesthetic: risk of aspiration and difficult airway due to bulbar dysfunction, central apnoea and respiratory sensitivity, autonomic instability with hypotension, management of blood glucose and desmopressin with strict fluid balance, glucocorticoid cover in corticotrophin deficiency and, in WS2, prior assessment of platelet function [1, 2, 38, 49]. Communication should be adapted to visual and hearing loss [49].

7.15. Agents and circumstances to avoid

Repeated noise exposure, increasing desmopressin without excluding bladder dysfunction, anticholinergics in urinary retention, bladder outflow obstruction and angle-closure glaucoma, and an indwelling catheter when there is an alternative are avoided [37, 60, 93]. The combination of SSRIs and desmopressin requires monitoring of serum sodium [83, 95]. Valproate should not be used outside a trial for Wolfram syndrome, and in the TREATWOLFRAM trial it required the pregnancy prevention programme [39, 98]. Antiplatelet agents and anticoagulants require careful consideration in WS2 [38].

7.16. Transition to adult care

Transition should begin at the start of the second decade, with assessment of understanding of the disease and of independent living skills, involving the young person and carers in a planned transfer [1]. The international consensus recommends the TRAQ (Transition Readiness Assessment Questionnaire) or Ready Steady Go tools [3]. The Italian consensus, without specific literature, recommends planning 1 to 2 years beforehand, with a reviewed individualised plan, collaboration and documentation between teams, transition supervised by the paediatrician or paediatric diabetologist and a review meeting after transfer, without a fixed age [2].

7.17. Investigational therapies and clinical trials

The aim of a specific therapy is to preserve endoplasmic reticulum function, calcium homeostasis, protein folding and redox balance in all tissues [2]. The approaches include chemical chaperones (4-phenylbutyric acid and tauroursodeoxycholic acid), reticulum calcium stabilisers (dantrolene), valproate as a neuroprotectant, GLP-1 agonists, iron chelators in WS2 and gene therapy [2, 40, 44, 99].

In the dantrolene trial, titration was 25 mg/day for 7 days with weekly doubling up to 200 mg/day in the adult, and 0.5 mg/kg/day (maximum 25 mg) with weekly doubling up to 2 mg/kg/day in the child; the final tolerated doses were 50 to 100 mg/day in the adult (mean 68.2 mg/day) and 0.5 to 2.0 mg/kg/day in the child (maximum 100 mg/day), close to those used in spasticity; the most common adverse effects attributed to the drug were mild fatigue and diarrhoea [21]. For comparison, the US label for Dantrium® in spasticity increases from 25 mg once daily to 100 mg three times daily in the adult (maximum 100 mg four times daily) and from 0.5 mg/kg once daily to 2 mg/kg three times daily in the child, with a boxed warning of hepatotoxicity and an instruction to stop if there is no benefit within 45 days [100]. In TREATWOLFRAM, valproate started at 10 mg/kg/day, in one or two doses, and was increased in increments of 200 or 400 mg over 4 weeks up to 40 mg/kg/day, with a maximum of 800 mg/day between 6 and 12 years and 1,600 mg/day from 12 years, in two doses 12 hours apart above 400 mg/day, with serial plasma levels and a mandatory pregnancy prevention programme [98]. In the WS2 pilot, 2 patients treated with deferiprone and N-acetylcysteine had a reported improvement in glycaemic control and correction of platelet aggregation [44]. In the Wfs1 knockout mouse, intravitreal injection of an adeno-associated viral vector AAV2 carrying human WFS1 at 30 days protected ganglion cells and axons and slowed the loss of acuity; there is no registered human trial [99]. Other preclinical data mention the sigma-1 receptor agonist PRE-084, which restores calcium transfer from the reticulum to mitochondria and reverses motor and cognitive deficits in mouse and zebrafish, and liraglutide, which reverses defects of β cells derived from induced pluripotent stem cells [42].

The PNDS notes that dantrolene hepatotoxicity, ranging from asymptomatic elevation of transaminases to fulminant hepatic failure, may occur even with sporadic and short-term use, that the fatal cases were associated with daily doses above 300 mg and that doses below 200 mg/day have been proposed as safe in patients without liver disease or other hepatotoxic drugs; the final doses in the Washington University trial remained below this threshold [21, 23]. The US label is less reassuring: it describes symptomatic hepatitis, fatal and non-fatal, at various doses, with a much lower incidence up to 400 mg/day than with 800 mg/day or more, a greater risk in women, after 35 years and with other drugs, and requires baseline and frequent transaminases during treatment, with no dose threshold below which the drug is safe [100].

The available trials have limitations that preclude firm conclusions. Only TREATWOLFRAM is randomised and controlled, and its results exist only in press releases, with an acknowledged baseline imbalance and without effect estimates with confidence intervals; the others are open-label, single-arm and with 11 to 21 patients, which does not allow the drug effect to be separated from the natural variation of outcomes such as C-peptide or visual acuity [19–21]. The results of different trials should not be compared indirectly with each other, because the populations, outcomes and durations differ. The dantrolene trial and the GLP-1 agonist cohort are negative for the efficacy outcomes, and exenatide was withdrawn owing to lack of recruitment [21, 55, 80]. An absence of published results is not equivalent to a negative result [98].

7.18. Developments in the last 12 months and level of evidence

The first TREATWOLFRAM results were released by the University of Birmingham on 23 June 2025 (republished in July 2025): vision worsened in all participants, at the same rate with valproate and placebo, and brainstem volume decreased in all, slightly more with valproate, with no unexpected safety problems; participants still on medication were advised to discuss stopping with their doctor [67]. The date falls outside the 12-month window, but it is the most recent status found.

7.19. Regulatory status

No drug has an approved indication for Wolfram syndrome in any jurisdiction; the drugs used are authorised for other conditions and prescribed symptomatically or off-label [55, 82, 101].

Sources: EMA, FDA (labels), Amylyx SEC report, TREATWOLFRAM protocol and Infomed SmPCs [78, 79, 82–85, 92–94, 98, 100, 101]. The FDA orphan designation for AMX0035 comes from the SEC report and that for valproate from the trial protocol; the FDA orphan designation database was not consulted directly. In Brazil, the ANVISA open database lists active registrations of Victoza® and Liraclick® (liraglutide), Dantrolen® (dantrolene) and the desmopressin presentations mentioned above, with no registration of idebenone [86].

7.20. Divergences between guidelines

Sources: [2, 3, 37]. The divergences in surveillance intervals are in section 11.

The 2019 French PNDS (HAS, OPHTARA reference centre) diverges from the other guidelines on the following points.

Sources: [1–3, 7, 23, 37, 38, 80, 83].

8. Reproductive Options and Family Counselling

8.1. Inheritance and recurrence risks by scenario

Counselling depends on the number of alleles and not only on the phenotype, and the same family may have both forms of inheritance [2, 9, 39].

Source: GeneReviews, genetic counselling [1]. The clinical course is variable even between siblings with the same variants and cannot be predicted from the type or location of the variants [1]. In the classic form, EURO-WABB indicates a recurrence risk of 25% for the parents and recommends informing affected adults and the extended family [37]. Most patients with a dominant form have a de novo variant; an apparently negative family history may be due to non-recognition, reduced penetrance, early death or late onset [1].

The French PNDS quantifies the risks to the offspring of a patient and of a healthy relative from the carrier frequency in the population, using Hardy-Weinberg equilibrium, and assumes that the members of the couple are not consanguineous [23]. Some French teams propose testing the partner of a carrier for the variant and, when a healthy relative makes a request, testing that relative first: if not a carrier, the couple's risk is nil or almost nil [23].

Source: PNDS, genetic counselling table [23]. RR is the factor of increase over the risk in the general population. In the last column, the PNDS prints 1/680 and 1/800 in the last two rows, values incompatible with the product of the carrier frequency by 1/4 and with the RR in the same table (415 and 450); the values presented here have been recalculated, and the PNDS text repeats the range of 1/460 to 1/800 [23]. With the lower prevalence estimates, the risk to the children of a patient falls below 1/800.

8.2. Heterozygous carriers

The most difficult point in counselling is the significance of a heterozygous variant in a relative of a recessive patient. GeneReviews found no conclusive evidence of increased risk in heterozygotes, but the existence of dominant forms, of variants that appear in both the recessive and the Wolfram-like form and of families with both forms shows that some carriers may have manifestations [1, 2, 9]. The international consensus asks that other causes be excluded before attributing a carrier's symptom to WFS1 [3]. In practice, it is useful to distinguish null variants, which cause recessive disease, from missense variants described as dominant, whose carrier has a risk of their own [17, 18]. The hypothesis of psychiatric predisposition in carriers comes from the studies by Swift and colleagues (1990, 1991 and 1998), who estimated the proportion of carriers in the US population at about 1% and the risk of psychiatric hospitalisation or suicide at about 8 times that of non-carriers [24]. These figures rest on methods predating or contemporary with the cloning of the gene, and a carrier frequency of 1% would imply, by Hardy-Weinberg equilibrium, a prevalence of about 1/40,000, far above the published population estimates (from about 1/770,000 to 1/100,000); they should therefore not be used in current counselling, and GeneReviews did not confirm an increased risk in heterozygotes [1, 24]. In the two Iranian families with severe disease due to homozygous p.(Asp211Asn) and p.(Gln486Ter), the heterozygous relatives had no manifestations [24]. The frequency of partners with isolated deafness in families with dominant forms, sometimes with GJB2 variants, justifies assessing the partner's hearing and, if indicated, genetics before estimating the risk to the offspring [25].

8.3. Cascade testing and minors

Testing of apparently asymptomatic at-risk relatives serves to identify early those who will benefit from treatment of the first manifestations (diabetes, optic atrophy, deafness) [1]. The Italian consensus recommends offering counselling and testing to parents, siblings, especially younger and asymptomatic ones, and other relatives, because WS, WFSL and the other WFS1-related disorders may coexist in the same family [2]. Pre-test counselling should address inheritance, sampling, consent, the appropriateness of the test and its sensitivity and specificity; post-test counselling includes the written report, interpretation, therapeutic options, follow-up, privacy and ownership of genetic information [2]. Testing an asymptomatic minor sibling for the family's biallelic variants has direct clinical benefit, because it allows multisystem surveillance to begin before the first manifestations [1, 2]. The sources consulted do not specifically address carrier testing in minors.

8.4. Prenatal diagnosis and preimplantation genetic testing

The optimal time to discuss the risk and prenatal or preimplantation testing is before pregnancy, and counselling should be offered to affected or at-risk young adults [1]. Prenatal diagnosis and preimplantation genetic testing are possible only after the family variants have been identified; EURO-WABB recommends discussing preimplantation testing with reference centres [1, 37]. Perspectives on prenatal testing vary, and the decision is personal [1].

In France, the indication is discussed in a specific genetics consultation and in a multidisciplinary prenatal diagnosis centre (CPDPN), which may offer termination of pregnancy for medical reasons if the fetus is affected, given the severity and the absence of a cure [23]. Prenatal diagnosis requires the proband's variants to have been identified and segregation confirmed in both parents, and is performed by chorionic villus sampling between 11 and 14 weeks of gestation or by amniocentesis from 16 weeks, with risks of fetal loss estimated by the PNDS at 1% and 0.5%, respectively, values predating the 2019 meta-analysis, which estimated the procedure-attributable risk at 0.20% for chorionic villus sampling (95% CI −0.13 to 0.52%) and 0.30% for amniocentesis (95% CI 0.11–0.49%), and at 0.12% or less when groups with similar chromosomal risk were compared [23, 102]; in preimplantation genetic testing, only embryos that do not carry both variants are transferred [23].

8.5. Fertility and reproduction in patients

Fertility is reduced, especially in men; the first report of male fertility in a severe form was published in 2013 [1, 47]. Gonadal surveillance from early adolescence allows fertility preservation to be offered before testicular function is lost, and the international consensus recommends timely counselling and preservation when subfertility is suspected [3, 62]. In women, the ability to conceive is generally retained, but ovarian insufficiency requiring oocyte donation and, in one case, absence of the uterus have been described [1, 49]. Pregnancy requires preconception glycaemic control and the planning described in subsection 7.13.

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and what each figure measures

The TREATWOLFRAM protocol uses an estimated prevalence of 1/500,000 [98]. Orphanet has no prevalence data for Wolfram-like syndrome and no separate entry for WS2, whose prevalence is included in that of Wolfram syndrome and is dominated by WFS1 [11, 38]. The international consensus considers the syndrome systematically underdiagnosed, and the Ecuadorian project found only 11% true type 1 diabetes in a cohort with early insulin-dependent diabetes, suggesting that many monogenic forms are classified as type 1 diabetes [3, 15].

9.2. Founder populations and ancestry

In the Valdivia project, in the province of Santa Elena (coast of Ecuador, an admixed population of European and Amerindian ancestry), 26 of 38 participants under 35 years with early insulin-dependent diabetes (mean onset at 3 years) had WS1 confirmed by exome; 23 were homozygous and all patients with WS1 carried the novel variant c.1135dup p.(Asp379GlyfsTer164), with a founder effect; about half were related through consanguinity; none had AVP deficiency and the neurological, visual and urinary complications were early [15]. In Sicily, the district of Messina has a prevalence about 14 times higher than the worldwide value [14, 39]. In Ashkenazi Jews, the p.(Arg558Cys) variant, associated with a mild form, is frequent [48]. In Palestinian families, the CISD2 variant c.109G>C has a carrier frequency of 1/40 [44]. The 1995 British cohort and that of Hardy 1999 included families of British, Pakistani and mixed Arab or African origin [1, 31]. In Spain, the duplication c.409_424dup (described as 425ins16) was the only variant identified in half of 16 families, homozygous in two thirds of them, and was attributed to a founder effect; in Italy, the deletion c.1362_1377del was the most frequent alteration, present in five unrelated families; in Lebanon, the complex allele WFS1(LIB) is always associated with the same haplotype and predominates in patients with isolated diabetes [5]. The diversity of private variants in most populations limits founder variant panels outside these groups [2].

9.3. United States and Europe

In the USA, the Washington University international registry and clinical study (NCT02841553, actively recruiting, 5,000 participants planned, uses the EURO-WABB criteria) is the main source of longitudinal data [54, 55]. In Europe, the French (96 patients with WFS1-related disorders), Italian, Spanish, German and British cohorts and the EURO-WABB registry (121 participants) underpin most of the evidence [37, 45, 56, 58].

9.4. Portugal and Brazil

In Portugal, the cross-sectional study from the Centro Hospitalar e Universitário de São João, in Porto, described 11 patients from 6 families with Wolfram syndrome and WFS1 variants: optic atrophy in 100%, diabetes in 90.9%, neurological, urological and psychiatric abnormalities in 63.6% each, sensorineural hearing loss in 54.5% and AVP deficiency in 9.1% [52]. Sequencing of the coding region and exon–intron junctions of WFS1 (NM_006005.3) showed a homozygous genotype in 5 patients and compound heterozygosity in 5; the eleventh had a single variant, p.(Asp797Tyr), de novo, which the authors interpret as dominant Wolfram-like syndrome [52]. The variants lay in exon 8 (84.6%) or exon 5 (15.4%) and included a novel variant, c.548T>C p.(Met183Thr), homozygous in two siblings and classified as of uncertain significance, so the molecular diagnosis of that family rests on a VUS [52]. No variant recurred between families apart from those shared by siblings, except the duplication c.2141_2164dup, present in two families, and the article does not describe copy number analysis [52]. It is a hospital cohort and not an estimate of national prevalence [52]. In Brazil, a four-generation family with 14 of 31 members affected by hereditary cataracts carried the WFS1 variant p.(Phe330_Phe331del), whose pathogenicity has not been clarified [50]. Brazil took part in the international consensus panel, with specialists from Porto Alegre (Casa dos Raros and Federal University of Rio Grande do Sul); Portugal does not appear in the panel's list of countries [3]. The sources consulted contain no Portuguese or Brazilian prevalence estimates, which does not mean that no published data exist.

9.5. Special populations

Pregnant women, patients with deafblindness and patients with cognitive impairment need adaptations in communication, care plans and decision support [2, 49]. The dominant neonatal form requires intensive neonatal and paediatric care and has its own prognosis [17]. In patients unable to self-catheterise because of ataxia, low vision or cognitive impairment, bladder management needs carers or surgical solutions [37].

9.6. European networks and patient associations

Endo-ERN (the European Reference Network on Rare Endocrine Conditions) includes the Association du syndrome de Wolfram (France) as a patient organisation and has organised webinars on diagnosis, biomarkers and therapeutic options; ERN-EYE has content on optic atrophy [22]. The Italian consensus lists the following associations and institutions [2]:

GeneReviews adds general resources on diabetes, deafness and blindness, and the Oxford Desk Reference mentions Diabetes UK and the Diabetes Genes website [1, 64].

In France, the syndrome is followed in the rare disease reference centres of the SENSGENE health network (sensory disorders), among them OPHTARA (Hôpital Necker-Enfants Malades and Hôpital Européen Georges-Pompidou, Paris), coordinator of the PNDS, MAOLYA (CHU de Montpellier, sensory disorders of genetic origin), CARGO (CHU de Strasbourg, ophthalmological disorders of genetic origin), the centre for congenital deafness of genetic origin (Necker) and the centre for retinal dystrophies (Hôpital des Quinze-Vingts), and of the FILNEMUS network, with the CALISSON mitochondrial disease centre (CHU de Nice); the PNDS also lists the visual investigation department of CHU de Lille [23]. The PNDS and the accompanying list of procedures and services serve as the basis for the care protocol in the application for exemption from patient co-payment (exonération du ticket modérateur) as a condition outside the list of long-term illnesses, and the attending physician guides the application for recognition of disability at the MDPH or MDA, with benefits such as the AEEH (allocation d'éducation de l'enfant handicapé) and the AJPP (allocation journalière de présence parentale) [23]. The syndrome may justify classification as a rare disability (handicap rare), with the support of the équipes relais handicaps rares (ERHR) in coordinating the care pathway; at school, the individualised reception protocol (PAI) and the personalised schooling plan (PPS) mobilise the CAMSP (0 to 6 years), the SAFEP (under 3 years), the SSEFS and support workers for pupils with disabilities (AESH), with the school doctor or occupational physician in follow-up, and the pathway should anticipate adult services [23]. The resources indicated are the Association du Syndrome de Wolfram (www.association-du-syndrome-de-wolfram.org), Orphanet, EURO-WABB and ARIBa (www.ariba-vision.org) [23].

10. Illustrative Clinical Case

10.1. Pregnancy in a woman with Wolfram syndrome type 1 treated with a hybrid closed-loop system

Caretto and colleagues described, in Italy, a woman with WS1 whose diabetes appeared at 11 years and was initially treated with diet alone [42]. At 32 years she had urinary urgency and enuresis, with detrusor overactivity treated with an anticholinergic; at 35 years she switched to an insulin pump with continuous glucose monitoring because of nocturnal hypoglycaemia, and later to an advanced hybrid closed-loop system (MiniMed 780G) [42]. Magnetic resonance imaging showed multisystem brain atrophy, and ophthalmological examination high myopia with nystagmus, ocular hypertension and optic atrophy [42]. The genetic diagnosis was made only at 39 years: compound heterozygosity for NM_006005.3:c.1060_1062del p.(Phe354del) and c.2534T>G p.(Ile845Ser), both in exon 8 [8, 42]. At 41 years she began preconception planning with an HbA1c of 6.8% (51 mmol/mol), discontinuation of the anticholinergic, folic acid 5 mg/day and genetic counselling of her partner [42]. After unsuccessful attempts at medically assisted reproduction, she conceived spontaneously at 44 years, with a preconception HbA1c of 7.2% and a body mass index of 18.6 kg/m² [42]. The closed-loop system was maintained outside the manufacturer's conditions of use, with an active insulin time of 2 hours and a target of 100 mg/dL; the total daily insulin dose was 25.6, 24.6 and 28.5 U in the three trimesters and 23.3 U post partum, and the glucose management indicator fell from 6.6% to 6.1% [42]. She received acetylsalicylic acid 100 mg/day from the first visit, had a urinary infection at 16 weeks and pre-eclampsia at 37+2 weeks (blood pressure up to 186/98 mmHg, ALT 85, AST 98, sFlt-1/PlGF ratio of 153, proteinuria of 0.686 g/24 h) [42]. An elective caesarean section was performed at 37+3 weeks, with the birth of a girl weighing 3,275 g, Apgar score 8/10, transient respiratory distress and neonatal hypoglycaemia at 2 hours [42]. Figure 1 of the article shows time in range, below range and above range during pregnancy and the postpartum period (https://pmc.ncbi.nlm.nih.gov/articles/PMC12450988/figure/F1/) [42].

The case illustrates several points discussed in this article: the delay of almost three decades between diabetes and molecular confirmation, adult-onset neurogenic bladder, the use of closed-loop technology to reduce hypoglycaemia and the course of pregnancy with pre-eclampsia and caesarean section, as in the published pregnancies [42]. Some practices are not routine: use of the closed-loop system in pregnancy outside the approved conditions and the folic acid dose of 5 mg/day are case-specific decisions and not recommendations of the guidelines for the syndrome, which do not address pregnancy in this detail [3, 42].

10.2. Complementary case: optic atrophy discovered on screening fundoscopy

Bazhar and colleagues described, in Morocco, a 20-year-old man with early bilateral deafness and insulin-dependent diabetes, referred for diabetic retinopathy screening and without visual complaints [68]. Acuity was 2/10 in both eyes, fundoscopy showed optic disc pallor, more marked in the temporal sector, without retinopathy (Figure 1: https://pmc.ncbi.nlm.nih.gov/articles/PMC12891919/figure/FIG1/), the visual fields had advanced defects (mean deviation of 21.1 dB in the left eye) and OCT showed nerve fibre layer thinning in all four quadrants (temporal 38 µm in the right eye and 29 µm in the left) [68]. Magnetic resonance imaging showed absence of the posterior pituitary bright spot and bilateral optic nerve atrophy (Figure 4: https://pmc.ncbi.nlm.nih.gov/articles/PMC12891919/figure/FIG4/), and AVP deficiency was confirmed [68]. The diagnosis was clinical: genetic testing was advised but not performed owing to financial constraints, and management included vitamin supplementation, low vision rehabilitation and follow-up [68]. The guidelines recommend molecular confirmation in all patients with suspected disease, and vitamin supplementation has no evidence in the syndrome [2, 3].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

In the 1995 British cohort, 10 of 45 patients had died at the time of the report; the causes were hypoglycaemic coma, status epilepticus, end-stage kidney disease due to recurrent urinary infections, central respiratory failure due to brainstem atrophy and suicide, and the median age at death was about 30 years (range 25 to 49) [1, 2, 16]. ISPAD likewise cites a median of 30 years, mainly due to neurodegenerative complications, and the Italian consensus gives a life expectancy of 30 to 40 years [2, 57]. Current survival is better than in the older series: in a national multidisciplinary clinic in the United Kingdom following 40 adults, the median age at the clinic was 37 years and the oldest patient was 65 years [1]. The prognosis depends mainly on neurological severity [2, 37]. A greater number of null alleles is associated with earlier onset, and the adult-onset forms and those with missense variants progress more slowly [3, 48, 53]. The dominant forms with optic atrophy and deafness appear to be non-progressive and without reduced life expectancy, with the exception of the neonatal form, whose first described patient died at 2 years [2, 17, 65]. In the review of 412 published patients by López de Heredia and colleagues, the median age of the 29 deceased patients was 27 years (± 11.4), with two peaks, at 24 and 45 years, and in the 2011 French series deaths were mainly due to respiratory failure or dysphagia due to brainstem involvement [24]. In WS2, reduced life expectancy has been described, without figures [38].

11.2. Surveillance

The following table uses the 2026 international consensus as its basis and indicates, where they differ, the intervals of EURO-WABB 2014, the 2019 French PNDS, the 2025 Italian consensus and GeneReviews.

In the non-classic form, GeneReviews recommends annual glycaemic monitoring until diabetes is diagnosed and quarterly thereafter, an annual eye examination, an annual audiogram with assessment of the need for hearing aids or implants, and annual social support and genetic counselling [1]. The GeneReviews on deafness places WFS1 in the group of progressive hearing losses and recommends audiological assessment at least annually, possibly six-monthly in children [60]. Transition to adult care begins at the start of the second decade (subsection 7.16) [1, 3].

11.3. Follow-up with treatment

Because no treatment modifies the disease, surveillance serves to detect and treat early the complications that determine mortality (hypoglycaemia, urinary infection and obstruction, central apnoea, aspiration, suicide) and to maintain independence with visual, auditory and motor rehabilitation [1, 16].

12. Summary and Final Considerations

Wolfram syndrome type 1 is an autosomal recessive neurodegenerative and endocrine disease caused by biallelic variants in WFS1, which encodes wolframin, an endoplasmic reticulum protein whose absence causes chronic reticulum stress and the death of β cells and neurons [1, 3]. The picture usually begins with non-autoimmune insulin-dependent diabetes, at about 6 years, followed by optic atrophy in the first decade; then, in variable proportions, arginine vasopressin deficiency, high-frequency hearing loss, neurogenic bladder, psychiatric disorders, hypogonadism and atrophy of the brainstem and cerebellum, which determines morbidity and mortality [2, 3, 16]. Wolfram syndrome type 2, caused by CISD2, overlaps clinically and adds peptic ulcer, bleeding and a platelet aggregation defect [1, 44].

The same WFS1 gene causes a group of dominant disorders, in which a single variant, often missense and dominantly acting, produces low-frequency hearing loss, congenital nuclear cataract, adult-onset diabetes, optic atrophy with early deafness or, at the severe extreme, neonatal diabetes with congenital deafness, cataracts and hypotonia [1, 17, 18]. The coexistence of recessive and dominant forms in the same family, and the uncertain significance of some heterozygous variants in carriers, make counselling dependent on careful classification of each variant on transcript NM_006005.3 [2, 8, 9].

The diagnosis should be suspected in any child with diabetes without autoantibodies and visual loss, in any infant with diabetes and deafness or cataracts, and in optic atrophy with early deafness; it is confirmed by a multigene panel with WFS1 and CISD2 and copy number analysis, bearing in mind that a negative result or a result with only one variant does not exclude the diagnosis [1, 3, 57]. Treatment is symptomatic and multidisciplinary: insulin with continuous monitoring and alarms adapted to visual loss, desmopressin with monitoring of serum sodium and the bladder, hormone replacement, visual and auditory rehabilitation, stepwise urological management, respiratory and swallowing care, mental health and educational and social support [2, 3, 37]. Annual or biennial surveillance of each system, according to the 2026 international consensus and the European guidelines, aims to treat early the complications that cause death [3, 37].

No disease-modifying therapy has demonstrated efficacy: dantrolene showed no efficacy, valproate did not slow visual loss in the controlled trial, GLP-1 agonists have contradictory observational data, and sodium phenylbutyrate–taurursodiol has open-label phase 2 results awaiting a phase 3 trial; ocular gene therapy is at the preclinical stage [19–21, 80, 99, 101]. The data from Portugal and Brazil in the sources consulted are limited to an ophthalmological cohort from Porto and a Brazilian family with cataracts, without prevalence estimates [50, 52].

13. References

1. Barrett T, Tranebjærg L, Gupta R, McCarthy L, Rendtorff ND, Williams D, Wright B, Dias R. WFS1 Spectrum Disorder. 2009 Feb 24 [updated 2022 Dec 1]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK4144/
2. Frontino G, Delvecchio M, Prudente S, Sordi VD, Barboni P, Di Giamberardino A, et al. SID/SIEDP expert consensus on optimizing clinical strategies for early detection and management of wolfram syndrome. J Endocrinol Invest. 2025;48(3):507-525. doi:10.1007/s40618-024-02495-z
3. Elliott J, Ahmadi S, Yu-Wai-Man P, Gladstone S, Gebel S, Lynch T, Barrett T, Urano F; International Wolfram Syndrome Clinical Guidelines Consortium. Consensus Recommendations for the Clinical Management of Wolfram syndrome Using a Delphi Method [preprint]. medRxiv. 2026 Jul 2. doi:10.64898/2026.07.02.26357130
4. NCBI Gene [Internet]. Bethesda (MD): National Center for Biotechnology Information. CISD2 (Gene ID 493856), WFS1 (Gene ID 7466) [cited 2026 Oct 26]. Available from: https://www.ncbi.nlm.nih.gov/gene/493856
5. OMIM® Online Mendelian Inheritance in Man. *606201 WOLFRAMIN ER TRANSMEMBRANE GLYCOPROTEIN; WFS1. Johns Hopkins University; last edited 23/06/2026. https://omim.org/entry/606201. Accessed 28 October 2026.
6. OMIM® Online Mendelian Inheritance in Man. #604928 WOLFRAM SYNDROME 2; WFS2. Johns Hopkins University; last edited 27/10/2023. https://omim.org/entry/604928. Accessed 28 October 2026.
7. OMIM® Online Mendelian Inheritance in Man. *611507 CDGSH IRON SULFUR DOMAIN PROTEIN 2; CISD2. Johns Hopkins University; last edited 30/07/2019. https://omim.org/entry/611507. Accessed 30 October 2026.
8. 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. Queried (GRCh38, MANE Select) on 2026 Oct 26: https://variantvalidator.org/
9. Lusk L, Black E, Vengoechea J. Segregation of two variants suggests the presence of autosomal dominant and recessive forms of WFS1-related disease within the same family: expanding the phenotypic spectrum of Wolfram Syndrome. J Med Genet. 2020;57(2):121-123. doi:10.1136/jmedgenet-2018-105782
10. Orphanet. Wolfram syndrome (ORPHA:3463) [Internet] [cited 2026 Oct 26]. Available from: https://www.orpha.net/en/disease/detail/3463
11. Orphanet. Wolfram-like syndrome (ORPHA:411590) [Internet] [cited 2026 Oct 26]. Available from: https://www.orpha.net/en/disease/detail/411590
12. Orphanet. Autosomal dominant non-syndromic genetic deafness (ORPHA:90635). Available from: https://www.orpha.net/en/disease/detail/90635
13. Orphanet. Early-onset nuclear cataract (ORPHA:98991). Available from: https://www.orpha.net/en/disease/detail/98991
14. Lombardo F, Salzano G, Di Bella C, Aversa T, Pugliatti F, Cara S, et al. Phenotypical and genotypical expression of Wolfram syndrome in 12 patients from a Sicilian district where this syndrome might not be so infrequent as generally expected. J Endocrinol Invest. 2014;37(2):195-202. doi:10.1007/s40618-013-0039-4
15. Palacio JE, Verdesoto MS, Bautista NL, Segura AJ, Cedeño RC, Flores NJ, et al. The Valdivia Project: genetic characterization of a unique Wolfram Syndrome cluster in coastal Ecuador. Orphanet J Rare Dis. 2026;21(1):173. doi:10.1186/s13023-026-04344-z
16. Barrett TG, Bundey SE, Macleod AF. Neurodegeneration and diabetes: UK nationwide study of Wolfram (DIDMOAD) syndrome. Lancet. 1995;346(8988):1458-1463. doi:10.1016/s0140-6736(95)92473-6
17. De Franco E, Flanagan SE, Yagi T, Abreu D, Mahadevan J, Johnson MB, et al. Dominant ER Stress-Inducing WFS1 Mutations Underlie a Genetic Syndrome of Neonatal/Infancy-Onset Diabetes, Congenital Sensorineural Deafness, and Congenital Cataracts. Diabetes. 2017;66(7):2044-2053. doi:10.2337/db16-1296
18. Roberts JP, Tang AF, Hernandez D, Carman B, Oiknine L, Brown C, et al. Natural history of 15 patients with autosomal dominant WFS1 pathogenic variants associated with sensorineural hearing loss and optic atrophy. Orphanet J Rare Dis. 2026;21(1):210. doi:10.1186/s13023-026-04348-9
19. The Snow Foundation. TREATWolfram — What's next? [Internet]. 2025 Nov 24 [cited 2026 Oct 26]. Available from: https://thesnowfoundation.org/treatwolfram-whats-next/
20. Urano F, Marshall BA, Hurst S, Robichaux-Viehoever A, Ahmadi S, Hershey T, et al. Phase II trial of sodium phenylbutyrate and taurursodiol in Wolfram syndrome. J Clin Invest. 2026;136(10):e198519. doi:10.1172/jci198519
21. Abreu D, Stone SI, Pearson TS, Bucelli RC, Simpson AN, Hurst S, et al. A phase Ib/IIa clinical trial of dantrolene sodium in patients with Wolfram syndrome. JCI Insight. 2021;6(15):145188. doi:10.1172/jci.insight.145188
22. European Reference Network on Rare Endocrine Conditions (Endo-ERN). Association du syndrome de Wolfram (ASW) — patient organisation [Internet] [cited 2026 Oct 26]. Available from: https://endo-ern.eu/patient-organisation/association-du-syndrome-de-wolfram-asw/
23. Haute Autorité de Santé; Centre de référence des maladies rares en ophtalmologie (OPHTARA). Protocole National de Diagnostic et de Soins (PNDS): Syndrome de Wolfram. Saint-Denis La Plaine: HAS; 2019 Oct 18 [cited 2026 Nov 6]. Available from: https://www.has-sante.fr/jcms/p_3113807
24. OMIM® Online Mendelian Inheritance in Man. #222300 WOLFRAM SYNDROME 1; WFS1. Johns Hopkins University; last edited 12/12/2025. https://omim.org/entry/222300. Accessed 28 October 2026.
25. OMIM® Online Mendelian Inheritance in Man. #614296 WOLFRAM-LIKE SYNDROME, AUTOSOMAL DOMINANT; WFSL. Johns Hopkins University; last edited 06/03/2024. https://omim.org/entry/614296. Accessed 28 October 2026.
26. OMIM® Online Mendelian Inheritance in Man. #600965 DEAFNESS, AUTOSOMAL DOMINANT 6; DFNA6. Johns Hopkins University; last edited 13/03/2023. https://omim.org/entry/600965. Accessed 28 October 2026.
27. OMIM® Online Mendelian Inheritance in Man. #116400 CATARACT 41; CTRCT41. Johns Hopkins University; last edited 29/06/2026. https://omim.org/entry/116400. Accessed 28 October 2026.
28. OMIM® Online Mendelian Inheritance in Man. #125853 TYPE 2 DIABETES MELLITUS; T2D. Johns Hopkins University; last edited 21/04/2025. https://omim.org/entry/125853. Accessed 28 October 2026.
29. Inoue H, Tanizawa Y, Wasson J, Behn P, Kalidas K, Bernal-Mizrachi E, et al. A gene encoding a transmembrane protein is mutated in patients with diabetes mellitus and optic atrophy (Wolfram syndrome). Nat Genet. 1998;20(2):143-148. doi:10.1038/2441
30. Strom TM, Hörtnagel K, Hofmann S, Gekeler F, Scharfe C, Rabl W, et al. Diabetes insipidus, diabetes mellitus, optic atrophy and deafness (DIDMOAD) caused by mutations in a novel gene (wolframin) coding for a predicted transmembrane protein. Hum Mol Genet. 1998;7(13):2021-2028. doi:10.1093/hmg/7.13.2021
31. Hardy C, Khanim F, Torres R, Scott-Brown M, Seller A, Poulton J, et al. Clinical and molecular genetic analysis of 19 Wolfram syndrome kindreds demonstrating a wide spectrum of mutations in WFS1. Am J Hum Genet. 1999;65(5):1279-1290. doi:10.1086/302609
32. Amr S, Heisey C, Zhang M, Xia XJ, Shows KH, Ajlouni K, et al. A homozygous mutation in a novel zinc-finger protein, ERIS, is responsible for Wolfram syndrome 2. Am J Hum Genet. 2007;81(4):673-683. doi:10.1086/520961
33. Eiberg H, Hansen L, Kjer B, Hansen T, Pedersen O, Bille M, et al. Autosomal dominant optic atrophy associated with hearing impairment and impaired glucose regulation caused by a missense mutation in the WFS1 gene. J Med Genet. 2006;43(5):435-440. doi:10.1136/jmg.2005.034892
34. Rendtorff ND, Lodahl M, Boulahbel H, Johansen IR, Pandya A, Welch KO, et al. Identification of p.A684V missense mutation in the WFS1 gene as a frequent cause of autosomal dominant optic atrophy and hearing impairment. Am J Med Genet A. 2011;155A(6):1298-1313. doi:10.1002/ajmg.a.33970
35. Bonnycastle LL, Chines PS, Hara T, Huyghe JR, Swift AJ, Heikinheimo P, et al. Autosomal dominant diabetes arising from a Wolfram syndrome 1 mutation. Diabetes. 2013;62(11):3943-3950. doi:10.2337/db13-0571
36. Berry V, Gregory-Evans C, Emmett W, Waseem N, Raby J, Prescott D, et al. Wolfram gene (WFS1) mutation causes autosomal dominant congenital nuclear cataract in humans. Eur J Hum Genet. 2013;21(12):1356-1360. doi:10.1038/ejhg.2013.52
37. Wolfram Syndrome Guideline Development Group (EURO-WABB). Management of Wolfram Syndrome: a clinical guideline. Wolfram Syndrome Clinical Management Guidelines, version 14 [Internet]. EURO-WABB; 2014 Apr 28 [cited 2026 Oct 26]. Available from: http://euro-wabb.org/media/downloads/guidelines/Wolfram_guideline_V14_%2028_04_2014.pdf (archived: http://web.archive.org/web/20220120063137/http://euro-wabb.org/media/downloads/guidelines/Wolfram_guideline_V14_%2028_04_2014.pdf)
38. GeneGuidance. CISD2 — clinical curation of the gene and Wolfram syndrome 2. Version of 2026-09-26.
39. GeneGuidance. WFS1 — clinical curation of the gene and associated conditions. Version of 2026-09-26.
40. Mishra R, Chen BS, Richa P, Yu-Wai-Man P. Wolfram syndrome: new pathophysiological insights and therapeutic strategies. Ther Adv Rare Dis. 2021;226330040211039518. doi:10.1177/26330040211039518
41. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 44, Disorders of Cellular Trafficking (44.1.3; Tables 44.1, 44.2).
42. Caretto A, Scarascia Mugnozza F, Valsecchi F, Pedone E, Pozzoni M, Rosa S, et al. Diabetes mellitus and pregnancy in Wolfram syndrome type 1: a case report with review of clinical and pathophysiological aspects. Front Med (Lausanne). 2025;12:1639884. doi:10.3389/fmed.2025.1639884
43. Samara A, Lugar HM, Hershey T, Shimony JS. Longitudinal Assessment of Neuroradiologic Features in Wolfram Syndrome. AJNR Am J Neuroradiol. 2020;41(12):2364-2369. doi:10.3174/ajnr.a6831
44. Cohen A, Karmi O, Abdulhag UN, Klopstock T, Sohn YS, Lobel O, et al. Reversible Mitochondrial Iron Toxicity in Wolfram Syndrome Type 2 Monogenic Diabetes. J Clin Endocrinol Metab. 2026;:dgag250. doi:10.1210/clinem/dgag250
45. Chaussenot A, Rouzier C, Quere M, Plutino M, Ait-El-Mkadem S, Bannwarth S, et al. Mutation update and uncommon phenotypes in a French cohort of 96 patients with WFS1-related disorders. Clin Genet. 2015;87(5):430-439. doi:10.1111/cge.12437
46. ClinVar [Internet]. Bethesda (MD): National Center for Biotechnology Information. WFS1 VCV000030556, VCV000004526, VCV000215413, VCV000290817, VCV001298954, VCV000092252, VCV000215378, VCV002203518, VCV000198835; CISD2 VCV000000892 [cited 2026 Oct 26]. Available from: https://www.ncbi.nlm.nih.gov/clinvar/
47. Haghighi A, Haghighi A, Setoodeh A, Saleh-Gohari N, Astuti D, Barrett TG. Identification of homozygous WFS1 mutations (p.Asp211Asn, p.Gln486*) causing severe Wolfram syndrome and first report of male fertility. Eur J Hum Genet. 2013;21(3):347-351. doi:10.1038/ejhg.2012.154
48. Oiknine L, Tang AF, Lee EM, Palaniappan N, Verma M, Vand K, et al. Genotype-based severity scoring system in Wolfram Syndrome: correlation with onset of cardinal symptoms and WFS1 gene variant types. Front Genet. 2026;171839135. doi:10.3389/fgene.2026.1839135
49. Esteban-Bueno G, Serrano Rodríguez ML. Pregnancy and Peripartum Multidisciplinary Management in Wolfram Syndrome Type 1: A Case Report. Diagnostics (Basel). 2026;16(8):1117. doi:10.3390/diagnostics16081117
50. Przysiezny B, Baccega TM, Neto EG, Khalil HH, Cordeiro LF, Taguchi FMC, et al. Hereditary cataract associated with a novel variant in WFS1. Ophthalmic Genet. 2026;47(5):512-516. doi:10.1080/13816810.2026.2655886
51. Rouzier C, Moore D, Delorme C, Lacas-Gervais S, Ait-El-Mkadem S, Fragaki K, et al. A novel CISD2 mutation associated with a classical Wolfram syndrome phenotype alters Ca2+ homeostasis and ER-mitochondria interactions. Hum Mol Genet. 2017;26(9):1599-1611. doi:10.1093/hmg/ddx060
52. Cunha AM, Estrela-Silva S, Barbosa-Breda J, Grangeia A, Carneiro Â, Falcão Dos Reis F, et al. Ophthalmic, Systemic and Genetic Features in Wolfram Syndrome. Neuroophthalmology. 2026;50(2):125-133. doi:10.1080/01658107.2025.2574723
53. de Heredia ML, Clèries R, Nunes V. Genotypic classification of patients with Wolfram syndrome: insights into the natural history of the disease and correlation with phenotype. Genet Med. 2013;15(7):497-506. doi:10.1038/gim.2012.180
54. Lee EM, Verma M, Palaniappan N, Pope EM, Lee S, Blacher L, et al. Genotype and clinical characteristics of patients with Wolfram syndrome and WFS1-related disorders. Front Genet. 2023;141198171. doi:10.3389/fgene.2023.1198171
55. ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine. Search: condition «Wolfram syndrome» (NCT02829268, NCT03717909, NCT04940572, NCT05676034, NCT05659368, NCT02882477, NCT01302327, NCT02841553, NCT02455414, NCT03951298, NCT07336966, NCT07313085, NCT07485413) [cited 2026 Oct 26]. Available from: https://clinicaltrials.gov/api/v2/studies?query.cond=Wolfram%20syndrome
56. Rigoli L, Aloi C, Salina A, Di Bella C, Salzano G, Caruso R, et al. Wolfram syndrome 1 in the Italian population: genotype-phenotype correlations. Pediatr Res. 2020;87(3):456-462. doi:10.1038/s41390-019-0487-4
57. Greeley SAW, Polak M, Njølstad PR, Barbetti F, Williams R, Castano L, et al. ISPAD Clinical Practice Consensus Guidelines 2022: The diagnosis and management of monogenic diabetes in children and adolescents. Pediatr Diabetes. 2022;23(8):1188-1211. doi:10.1111/pedi.13426
58. Rohayem J, Ehlers C, Wiedemann B, Holl R, Oexle K, Kordonouri O, et al. Diabetes and neurodegeneration in Wolfram syndrome: a multicenter study of phenotype and genotype. Diabetes Care. 2011;34(7):1503-1510. doi:10.2337/dc10-1937
59. Karzon R, Narayanan A, Chen L, Lieu JEC, Hershey T. Longitudinal hearing loss in Wolfram syndrome. Orphanet J Rare Dis. 2018;13(1):102. doi:10.1186/s13023-018-0852-0
60. Shearer AE, Hildebrand MS, Odell AM, Smith RJH. Genetic Hearing Loss Overview. 1999 Feb 14 [updated 2023 Apr 6]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1434/
61. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 122–124 (Deafness in early childhood).
62. Rohayem J, Cunningham O, Williams D, Wistuba J, McCarthy L, Barrett TG, et al. Gonadal function in males with WFS1 spectrum disorder (Wolfram syndrome)-A European cohort perspective. Andrology. 2026;14(2):398-410. doi:10.1111/andr.70049
63. Wragg R, Dias RP, Barrett T, McCarthy L. Bladder dysfunction in Wolfram syndrome is highly prevalent and progresses to megacystis. J Pediatr Surg. 2018;53(2):321-325. doi:10.1016/j.jpedsurg.2017.11.025
64. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 391–393 (Diabetes mellitus).
65. de Muijnck C, Brink JBT, Bergen AA, Boon CJF, van Genderen MM. Delineating Wolfram-like syndrome: A systematic review and discussion of the WFS1-associated disease spectrum. Surv Ophthalmol. 2023;68(4):641-654. doi:10.1016/j.survophthal.2023.01.012
66. Mozzillo E, Delvecchio M, Carella M, Grandone E, Palumbo P, Salina A, et al. A novel CISD2 intragenic deletion, optic neuropathy and platelet aggregation defect in Wolfram syndrome type 2. BMC Med Genet. 2014;15:88. doi:10.1186/1471-2350-15-88
67. The Snow Foundation. TREATWOLFRAM international clinical trial of sodium valproate in Wolfram syndrome: interim results [Internet]. 2025 Jul 20 (republication of the University of Birmingham press release of 2025 Jun 23) [cited 2026 Oct 26]. Available from: https://thesnowfoundation.org/treatwolfram-international-clinical-trial-of-sodium-valproate-in-wolfram-syndrome-interim-results/
68. Bazhar H, Loubna M, Bouslous N, Omar M. Routine Fundoscopy Uncovering Wolfram Syndrome in a Diabetic Patient: A Case Report. Cureus. 2026;18(1):e101313. doi:10.7759/cureus.101313
69. Urano F. Wolfram Syndrome: Diagnosis, Management, and Treatment. Curr Diab Rep. 2016;16(1):6. doi:10.1007/s11892-015-0702-6
70. Khanim F, Kirk J, Latif F, Barrett TG. WFS1/wolframin mutations, Wolfram syndrome, and associated diseases. Hum Mutat. 2001;17(5):357-367. doi:10.1002/humu.1110
71. OMIM® Online Mendelian Inheritance in Man. #249270 THIAMINE-RESPONSIVE MEGALOBLASTIC ANEMIA SYNDROME; TRMA. Johns Hopkins University; last edited 09/12/2022. https://omim.org/entry/249270. Accessed 30 October 2026.
72. OMIM® Online Mendelian Inheritance in Man. #125250 OPTIC ATROPHY PLUS SYNDROME. Johns Hopkins University; last edited 02/07/2026. https://omim.org/entry/125250. Accessed 30 October 2026.
73. De León DD, Pinney SE. Permanent Neonatal Diabetes Mellitus. 2008 Feb 8 [updated 2024 Nov 14]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1447/
74. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 10, Disorders of Oxidative Phosphorylation (10.1, Table 10.2).
75. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, Clinical Approach to Inborn Errors of Metabolism in Paediatrics (1.1.1.2; 1.2.2; 1.5; 1.6.3, Table 1.35).
76. Mylan Pharmaceuticals Inc. Glucagon for injection. Prescribing information. ANDA204468. Effective 2024-12-15 [cited 2026 Oct 26]. DailyMed set ID 15b49500-0aae-4e25-9a1d-709983233cc6. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=15b49500-0aae-4e25-9a1d-709983233cc6
77. Frontino G, Raouf T, Canarutto D, Tirelli E, Di Tonno R, Rigamonti A, et al. Case Report: Off-Label Liraglutide Use in Children With Wolfram Syndrome Type 1: Extensive Characterization of Four Patients. Front Pediatr. 2021;9755365. doi:10.3389/fped.2021.755365
78. Novo Nordisk. Victoza (liraglutide) injection. Prescribing information. NDA022341. Revised 2025-10 [cited 2026 Oct 26]. DailyMed set ID 5a9ef4ea-c76a-4d34-a604-27c5b505f5a4. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=5a9ef4ea-c76a-4d34-a604-27c5b505f5a4
79. European Medicines Agency. Victoza (liraglutide): EPAR [Internet]. Authorised 2009-06-30 [cited 2026 Oct 26]. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/victoza
80. Lee LA, Tang AF, Asako AE, Ning SF, Reed HA, Duncan E, et al. Evaluating the use of GLP-1 receptor agonists in Wolfram syndrome patients. Front Endocrinol (Lausanne). 2026;171847818. doi:10.3389/fendo.2026.1847818
81. L'Amie A, Vincent AL, Jefferies CA, Braatvedt G, Wiltshire E, Mouat F, et al. Wolfram syndrome and diabetes mellitus in Aotearoa, New Zealand: Phenotype and response to GLP-1 receptor agonist therapy. Diabet Med. 2026;43(8):e70405. doi:10.1111/dme.70405
82. European Medicines Agency. Orphan designations for Wolfram syndrome: EU/3/14/1428 (sodium valproate), EU/3/16/1800 (dantrolene sodium, withdrawn), EU/3/22/2667 (liraglutide), EU/3/24/2954 (sodium phenylbutyrate/ursodoxicoltaurine) [Internet; cited 2026 Oct 26]. Available from: https://www.ema.europa.eu/en/medicines/human/orphan-designations/eu-3-24-2954
83. Ferring. Minirin 0,12 mg liofilizado oral (desmopressina): Resumo das Características do Medicamento [Internet]. Lisboa: INFARMED, Infomed (med_id 38488) [cited 2026 Oct 26]. Available from: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=38488&tipo_doc=rcm
84. Ferring Pharmaceuticals Inc. DDAVP (desmopressin acetate) tablets. Prescribing information. NDA019955. Revised 2021-02 [cited 2026 Oct 26]. DailyMed set ID 6d55baa9-2b62-469c-93ae-3909ab249332. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=6d55baa9-2b62-469c-93ae-3909ab249332
85. Oceanside Pharmaceuticals. Desmopressin acetate nasal spray. Prescribing information. ANDA074830. Revised 2025-06 [cited 2026 Oct 26]. DailyMed set ID 8034989a-45f5-436b-8b1f-42108ee2a21a. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=8034989a-45f5-436b-8b1f-42108ee2a21a
86. Agência Nacional de Vigilância Sanitária (ANVISA). Dados abertos — medicamentos registados (DADOS_ABERTOS_MEDICAMENTOS.csv) [Internet] [cited 2026 Oct 26]. Available from: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
87. Endo USA. Aveed (testosterone undecanoate) injection. Prescribing information. NDA022219. Effective 2025-07-15 [cited 2026 Nov 1]. DailyMed set ID f80f025b-17d8-40af-8739-20ce07902045. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=f80f025b-17d8-40af-8739-20ce07902045
88. Prasco Laboratories. Testosterone gel 1%. Prescribing information. NDA021015. Effective 2024-12-01 [cited 2026 Nov 1]. DailyMed set ID 68b2375d-0365-4797-bc3d-63531c9f42fd. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=68b2375d-0365-4797-bc3d-63531c9f42fd
89. AbbVie Inc. Synthroid (levothyroxine sodium) — prescribing information [Internet]. Revised 2024-02-20 (NDA021402). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=1e11ad30-1041-4520-10b0-8f9d30d30fcc
90. Laboratórios Pfizer. Genotropin 5,3 mg e 12 mg pó e solvente para solução injetável (somatropina): Resumo das Características do Medicamento [Internet]. Lisboa: INFARMED, Infomed (med_id 3847) [cited 2026 Oct 26]. Available from: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=3847&tipo_doc=rcm
91. Pfizer (ProPharma Distribution). Solu-Cortef (hydrocortisone sodium succinate) for injection. Prescribing information. NDA009866. Effective 2025-07-18 [cited 2026 Oct 26]. DailyMed set ID 386942d3-c75f-25f1-e063-6394a90a5af2. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=386942d3-c75f-25f1-e063-6394a90a5af2
92. European Medicines Agency. Raxone (idebenone): EPAR – product information [Internet]. Authorised 2015-09-08 [cited 2026 Oct 26]. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/raxone
93. Cheplapharm Arzneimittel GmbH. Ditropan 5 mg comprimidos (cloridrato de oxibutinina): Resumo das Características do Medicamento [Internet]. Lisboa: INFARMED, Infomed; text revised March 2024 [cited 2026 Oct 26]. Available from: https://extranet.infarmed.pt/INFOMED-fo/
94. Aphena Pharma Solutions. Oxybutynin chloride extended-release tablets. Prescribing information. ANDA078503. Revised 2023-05 [cited 2026 Oct 26]. DailyMed set ID 02ff3be7-fc5c-4b91-8c67-ecdf6e19c42a. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=02ff3be7-fc5c-4b91-8c67-ecdf6e19c42a
95. Dista Products Company (Eli Lilly). Prozac (fluoxetine) capsules. Prescribing information. NDA018936. Effective 2026-01-08 [cited 2026 Oct 26]. DailyMed set ID c88f33ed-6dfb-4c5e-bc01-d8e36dd97299. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=c88f33ed-6dfb-4c5e-bc01-d8e36dd97299
96. European Medicines Agency. Prozac and associated names (fluoxetine): Article 6(12) referral — Annex I–III (CHMP/46089/06); CHMP opinion of 1 Jun 2006 and European Commission decision of 21 Aug 2006 [accessed 7 Nov 2026]. Available from: https://www.ema.europa.eu/en/medicines/human/referrals/prozac
97. Allergan Inc. Celexa (citalopram) tablets. Prescribing information. NDA020822. Effective 2023-10-09 [cited 2026 Oct 26]. DailyMed set ID 4259d9b1-de34-43a4-85a8-41dd214e9177. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=4259d9b1-de34-43a4-85a8-41dd214e9177
98. Dias RP, Brock K, Hu K, Gupta R, Martin U, Peet A, et al. Sodium valproate, a potential repurposed treatment for the neurodegeneration in Wolfram syndrome (TREATWOLFRAM): trial protocol for a pivotal multicentre, randomised double-blind controlled trial. BMJ Open. 2025;15(2):e091495. doi:10.1136/bmjopen-2024-091495
99. Jagodzinska J, Péquignot M, Sarzi E, Quiles M, Cazevieille C, Müller A, et al. WFS1 gene delivery rescues visual function in a mouse model of Wolfram syndrome. Acta Neuropathol Commun. 2026;14(1):120. doi:10.1186/s40478-026-02295-x
100. Par Health USA. Dantrium (dantrolene sodium) capsules. Prescribing information. NDA017443. Revised 2026-08 [cited 2026 Oct 26]. DailyMed set ID 57b4eec1-49e1-432c-95ae-0fe36e32d64b. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=57b4eec1-49e1-432c-95ae-0fe36e32d64b
101. Amylyx Pharmaceuticals Inc. Form 10-Q for the quarterly period ended June 30, 2026 [Internet]. US Securities and Exchange Commission; 2026 [cited 2026 Oct 26]. Available from: https://www.sec.gov/Archives/edgar/data/0001658551/000119312526336682/amlx-20260630.htm
102. Salomon LJ, Sotiriadis A, Wulff CB, Odibo A, Akolekar R. Risk of miscarriage following amniocentesis or chorionic villus sampling: systematic review of literature and updated meta-analysis. Ultrasound Obstet Gynecol. 2019;54(4):442–51. doi:10.1002/uog.20353. PMID: 31124209.
103. Ganie MA, Laway BA, Nisar S, Wani MM, Khurana ML, Ahmad F, et al. Presentation and clinical course of Wolfram (DIDMOAD) syndrome from North India. Diabet Med. 2011;28(11):1337-1342. doi:10.1111/j.1464-5491.2011.03377.x