Canavan Disease

Autosomal recessive spongiform leukodystrophy due to aspartoacylase (ASPA) deficiency, with accumulation of N-acetylaspartate in the brain and body fluids; spectrum ranging from the typical infantile form to the mild or juvenile form.

Literature and regulatory information updated: 26 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Historical overview

Canavan disease was first identified as a neuropathological entity, spongy degeneration of the brain, and the designation "Van Bogaert and Bertrand type" that GeneReviews retains among the synonyms recalls this morphological origin [1]. The neuropathology of the typical form is that of a subcortical spongy degeneration, and electron microscopy shows swollen astrocytes and distorted mitochondria [1]. The original neuropathological descriptions were not consulted directly for this article, so the dates and authors of those first publications are not given. For decades diagnosis relied on autopsy and brain biopsy, and survival described in the classic series, such as that of Lyon and colleagues, was limited to about 3 years, with rare exceptions reaching 10 years [10].

Before the enzyme defect was identified, patients with N-acetylaspartic aciduria and leukodystrophy or progressive cerebral atrophy had been described, with deficient aspartoacylase activity in one case (Hagenfeldt, 1987) and normal activity in another (Kvittingen, 1986), as well as two siblings with N-acetylaspartic aciduria, macrocephaly and leukodystrophy (Divry, 1988), without these findings being linked to Canavan disease [6]. In the US, the disease was observed in infants of Jewish ancestry whose forebears had lived in Vilna (Banker and colleagues, 1964) [6]. The classic OMIM synopsis describes onset in early infancy, atony of the neck muscles, hypotonia, hyperextension of the legs with flexion of the arms, blindness, severe intellectual disability, megalocephaly and death at a mean age of 18 months, a historical figure that recent survival data exceed (Section 3.11) [6].

The biochemical milestone came in 1988, when Matalon and colleagues demonstrated, in three patients from two families with spongiform degeneration, increased urinary excretion and plasma concentration of N-acetylaspartic acid and profound aspartoacylase deficiency in cultured skin fibroblasts [17]. In the same work the authors noted that the enzyme is active in amniocytes and chorionic villi, which opened the way to enzymatic prenatal diagnosis [17]. In an addendum, the same authors described the deficiency in a third case, from an Australian family, and in 1989 they published 21 patients, 14 with confirmation by brain biopsy, all with urinary NAA excretion about 200 times that of age-matched controls and obligate carriers and with aspartoacylase deficiency [6]. The metabolic designation N-acetylaspartic aciduria stems from this discovery, and Saudubray includes the disease in the group of cerebral organic acidurias, alongside glutaric aciduria type 1, the 2-hydroxyglutaric acidurias and succinic semialdehyde dehydrogenase deficiency [2].

In 1993, Kaul and colleagues cloned the human aspartoacylase cDNA, of 1435 base pairs, demonstrated the activity of the protein expressed in bacteria and identified the A>C substitution at nucleotide 854 (Glu285Ala; historical numbering, = NM_000049.4:c.854A>C p.(Glu285Ala)) in 85% of 34 patient alleles, at a position assigned to the catalytic domain [18]. By 1993 the same group had already diagnosed 145 patients at its centre, suggesting a disease more prevalent than previously thought [6]. The following year, analysis of 128 chromosomes from 64 unrelated probands defined the two Ashkenazi founder variants, the splice variant c.433-2A>G and the p.Ala305Glu variant, predominant in non-Jewish patients of European ancestry [19]. These data enabled carrier screening programmes directed at the Ashkenazi Jewish population, guided in the US by the 2008 technical standard of the American College of Medical Genetics and Genomics (ACMG) [2, 20].

The first gene transfer programme into the mammalian central nervous system (CNS) directed at the disease was published in 2000, followed by the clinical protocol for gene therapy with an adeno-associated virus (AAV) serotype 2 vector administered neurosurgically [21, 22]. In the 2020s, research advanced in three directions: prospective characterisation of the natural history [10, 23], identification by long-read sequencing of a retrotransposon insertion in intron 4 of ASPA, which became the most frequent known pathogenic variant [9, 12], and gene therapy trials with new-generation vectors, with the first phase 1/2 publication in 2025 [13].

1.2. Nomenclature and definition of the entities

GeneReviews distinguishes the typical form, also called infantile, from the atypical form, also called juvenile [1]. Orphanet uses the designations severe form (ORPHA:314911), defined as a rapidly progressive neurodegenerative disease with leukodystrophy, macrocephaly, severe developmental delay and hypotonia, and mild form (ORPHA:314918), defined by mild speech or motor developmental delay [5]. Saudubray further describes two rare clinical variants within the spectrum, a congenital one, with symptoms at birth or shortly after, and a juvenile one, with onset after 5 years [2]. All these forms result from biallelic variants in the same gene and constitute a continuum of severity related to residual enzyme activity and to urinary NAA concentration, and not a class of genetically distinct syndromes [9, 23, 24].

The typical form is defined by onset in the first months of life, with hypotonia, lack of head control and macrocephaly, acquisition of few milestones and regression [1]. The atypical or mild form is defined by developmental delay in the first years, sometimes without regression, normal or increased head circumference, nystagmus or visual impairment and ataxia or impaired coordination, with a smaller rise in urinary NAA [1, 23]. In practice, the boundary between the two forms is set by the milestones achieved and the trajectory, because no published consensus clinical criteria exist [1].

1.3. Scope of the article

The article covers the typical form, the mild or juvenile form and the gene therapies in trials. The general approach to leukodystrophies, pattern-based imaging algorithms and the classification of genetic leukoencephalopathies are not developed; GM2 gangliosidoses, glutaric acidaemia type 1, Leigh syndrome and mitochondrial disorders, metachromatic leukodystrophy, Krabbe disease and X-linked adrenoleukodystrophy are included in the differential diagnosis. Hypoacetylaspartia due to NAT8L deficiency and defects of the malate-aspartate shuttle are covered in the differential diagnosis.

2. Molecular Genetics and Pathophysiology

2.1. The ASPA gene and aspartoacylase

The ASPA gene is located at 17p13.2 and encodes aspartoacylase, also called aminoacylase 2, a hydrolase that cleaves N-acetyl-L-aspartate into L-aspartate and acetate [1, 2]. The gene has 6 exons and encodes a protein of 313 amino acids, of about 36 kDa, with 92% identity to the bovine protein, one potential N-glycosylation site and five phosphorylation sites; by Northern blot, a 1.44 kb transcript is detected in the liver and an additional 5.4 kb transcript in other tissues, with maximal expression in skeletal muscle, followed by kidney and brain [3]. The MANE reference transcript is NM_000049.4, which encodes the protein NP_000040.1; the numbering of the coding nucleotides coincides with that of transcript NM_000049.2 used in the GeneReviews variant table [1, 4, 12]. The Glu285Ala substitution, described when the gene was cloned, lies at a residue assigned to the Ser-His-Glu catalytic triad [18]. The enzyme is expressed in many tissues and, outside the CNS, acts as a scavenger of NAA from body fluids; its absence, however, produces selective accumulation of NAA in the brain and disease limited to the white matter [1]. The disease mechanism is loss of function [1]. Aminoacylase 1 (ACY1), a distinct enzyme, cleaves acylated L-amino acids, with the exception of L-aspartate [3].

2.2. N-acetylaspartate metabolism and cellular compartmentalisation

NAA is a compound characteristic of the brain, at a concentration of the order of 20 mM, synthesised in neurons by L-aspartate N-acetyltransferase (NAT8L) from L-aspartate and acetyl-coenzyme A [2]. In the brain, aspartoacylase is located in oligodendrocytes, so NAA has to pass from neurons to oligodendrocytes to be hydrolysed [2]. This compartmentalisation is interpreted as a system channelling neuronal acetate to oligodendrocytes, where it is a main substrate for the synthesis of myelin lipids [2]. NAA metabolism is also coupled to the malate-aspartate shuttle (the AGC1 carrier, encoded by SLC25A12), which governs the redox state between cytoplasm and mitochondria [2, 25].

There are two main pathophysiological hypotheses, which are not mutually exclusive. In the acetate deficiency hypothesis, the inability to hydrolyse NAA reduces brain acetate and myelin lipid synthesis; in the aspartoacylase-deficient mouse, a 30% reduction in total myelin lipids and a marked fall in brain acetate were observed at the peak of postnatal myelination, with normal acetate in liver and kidney, and the white matter of a patient homozygous for p.Glu285Ala had a significant reduction in the complex glycolipids that co-migrate with cerebroside and sulfatide [2, 3, 26]. In the osmotic hypothesis, NAA would act as a molecular water pump, carrying metabolic water from glucose oxidation from neurons to oligodendrocytes; lack of catabolism would cause osmotic dysregulation and the spongy appearance of the white matter [2]. The cytotoxic oedema with restricted diffusion seen in early stages of the disease is compatible with an active postnatal injurious process, although diffuse oedema with increased brain water also occurs in diseases without NAA accumulation, which led Merrill and colleagues to allow that NAA may not be the primary cause of the oedema [27]. The cerebral organic acidurias also share the hypothesis that the accumulated compounds interfere with glutamatergic and GABAergic neurotransmission, energy metabolism and myelin metabolism [2].

Animal models have added a third element. In an optimised AAV gene replacement therapy in deficient mice, correction of ASPA expression in astrocytes, independently of oligodendrocytes, was sufficient to rescue the lethal phenotype early, completely and durably, pointing to an astrocytic metabolic "sink" for NAA as a therapeutic mechanism [28]. In a mouse model, N-acetylaspartylglutamate (NAAG) synthetase deficiency had little influence on pathogenesis [29], and knockout of the sodium-dependent dicarboxylate transporter Slc13a3 normalised brain NAA, improved motor performance and prevented cerebellar and thalamic vacuolation in the Canavan disease mouse, with a partial effect in heterozygosity (Wang and colleagues, 2021); the conditional astroglial version also had a therapeutic effect [3, 30].

2.3. Pattern of inheritance

Inheritance is autosomal recessive. The disease requires biallelic pathogenic or likely pathogenic variants, in homozygosity or compound heterozygosity, and heterozygous carriers are asymptomatic [1]. When only one parent is a carrier, with confirmed parentage, the possible explanations are a de novo variant in the proband, postzygotic mosaicism in a parent or uniparental isodisomy [1]. A heterozygous 17p13.3 microdeletion may also unmask a recessive variant on the other allele [31].

2.4. Variant spectrum and nomenclature

More than 100 pathogenic ASPA variants have been submitted to the Leiden Open Variation Database (LOVD), ClinVar and the Human Gene Mutation Database (HGMD) [12]. According to the professional HGMD data compiled by GeneReviews, about 99% of the pathogenic variants described are detectable by sequencing and about 1% are deletions or duplications of exons or of the whole gene [1, 32]. This proportion refers to the set of known variants and not to the clinical sensitivity of each method, which is lower for short-read sequencing than these figures suggest, because the most frequent variant is an intronic insertion that this technology does not identify (Section 2.5) [9, 12]. In non-Jewish patients the variants are diverse and largely private [2]. Shaag and colleagues identified, in 19 non-Jewish European patients, nine variants (four point variants, four deletions and one exon skip), all in exons 4 to 6, detecting 31 of the 38 alleles (81.6%); Kaul and colleagues described in 1996 eight novel variants in 25 non-Jewish patients, which together with three already known explained 40 of the 50 mutant chromosomes, and all caused aspartoacylase deficiency on in vitro expression; Zeng and colleagues found 24 different variants, 14 novel, in 22 non-Jewish patients [3, 6, 33].

The following table brings together the variants with population significance or phenotypic correlation, including the 12 allelic variants selected by OMIM, normalised in VariantValidator on the MANE transcript and the GRCh38 genome; all validated without discordance warnings, and the OMIM variants described in historical cDNA nomenclature (for example, 854A-C, 693C-A, 4-bp del 876AGAA, deletion of T at position 32) keep the same numbering on the current transcript [3, 4].

GeneReviews is internally inconsistent regarding p.Ala305Glu: the text describes very low enzyme activity and the variant table indicates complete loss [1]. In Kaul's expression studies, p.Tyr231Ter and p.Ala305Glu abolished activity and p.Glu285Ala retained 2.5% [19]. The observation of less severely affected patients with genotypes that include p.Ala305Glu, such as p.[Glu285Ala];[Ala305Glu], p.[Ala305Glu];[Ala305Glu] and p.[Ala287Thr];[Ala305Glu], coexists with typical cases homozygous for the same variant, such as the clinical case in Section 10 [10, 27]. OMIM also records expression of p.Ala305Glu and p.Tyr231Ter in COS-1 cells with complete loss of activity and homozygosity for p.Ala305Glu in patients with severe and mild forms (Shaag, 1995) [3]. The safest reading is that p.Ala305Glu has very low or absent residual activity and does not by itself predict the clinical form.

2.5. The SVA_E retrotransposon insertion in intron 4

In 2025, Dominguez Gonzalez and colleagues studied eight individuals from seven families, of European, European-Uruguayan and European-American ancestry, with a clinical, biochemical and imaging diagnosis of Canavan disease and short-read sequencing that was negative or showed a single allele [12]. Long-read sequencing (Oxford Nanopore and PacBio HiFi) identified an insertion of about 2600 bp of an SVA_E retrotransposon in intron 4 of ASPA, homozygous in three individuals and in trans with a known pathogenic variant in five (c.914C>A in three, c.634+1G>T in one and c.820G>A p.(Gly274Arg) in another) [12]. RNA sequencing of fibroblasts, with and without cycloheximide, showed that the insertion creates a new splice acceptor site, with inclusion of an exon containing part of the SVA_E, termination of transcription within the insertion and nonsense-mediated decay of the transcript [12].

The population frequency explains the clinical importance of the finding. In the Genome Aggregation Database (gnomAD) v4, in about 126,000 haplotypes, the insertion was observed in 66 heterozygotes and no homozygotes, with an allele frequency of 0.0005 (1/1909 haplotypes), higher than that of the c.914C>A variant, which is 0.0003 (1/3314) [12]. The maximum allele frequency compatible with pathogenicity, calculated for an assumed prevalence of 1/10,000 with complete penetrance, is 0.0013, so the observed value does not contradict the classification [12]. The insertion is present in all gnomAD ancestry groups except the Middle Eastern group, which has fewer than 100 haplotypes represented, and four individuals share a haplotype with more than 99% identity over a region of about 2.1 Mb, indicating a single, ancient insertion event predating population divergence, and not a recent founder effect [12]. The variant was classified as pathogenic under the ACMG and Association for Molecular Pathology (ACMG/AMP) criteria (PM3_VeryStrong and PS3) and deposited in ClinVar (SCV005407762) [12, 36].

Independent confirmation came from the Hamburg group. In a cohort of more than 55 patients, about 13% remained without a complete molecular diagnosis, with no or only one pathogenic variant [9]. Long-read sequencing after CRISPR/Cas9 enrichment of a region of more than 38 kb identified in the five patients studied the same heterozygous insertion of about 2630 bp, described as NC_000017.10:g.3393511_3393512insSVA_E, in cis with the intronic variant c.236+1187A>C (gnomAD frequency 0.00887, with no predicted effect on splicing), which acts as a haplotype marker [9]. The authors developed a conventional polymerase chain reaction (PCR) that distinguishes a 260 bp product from the normal allele from a product of about 3000 bp from the allele with the insertion [9].

2.6. Large deletions and structural variants

Deletions and duplications account for about 1% of the variants described [1]. Zeng and colleagues described three novel large deletions in non-Jewish patients [37], Caliebe and colleagues a homozygous 439 kb deletion at 17p13.3 identified by single-nucleotide polymorphism microarray (SNP array) [38], and Cozzolino and colleagues a heterozygous 17p13.3 microdeletion that unmasked a recessive variant on the other allele [31]. Copy-number analysis targeted to ASPA therefore has a place when sequencing identifies only one variant [1].

2.7. Genotype–phenotype correlation

Severity correlates with residual aspartoacylase activity. In a functional study of 12 novel missense variants in 14 patients, with expression in HEK293 cells and measurement of the aspartate released by liquid chromatography with tandem mass spectrometry (LC-MS/MS), the variants of the four patients with the mild or atypical form were those with the highest residual activity [24]. The p.Tyr231Ter and p.Glu285Ala variants, in homozygosity or compound heterozygosity, are associated with the typical form [1]. The p.Arg71His, p.Asp204His, p.Pro257Arg and p.Tyr288Cys variants are associated with the atypical form even in compound heterozygosity with a null allele [1]. In the natural history series of Nagy and colleagues, p.Arg71His and p.Tyr288Cys occurred in 56% (14/25) of patients with a mild phenotype and in none (0/39) of patients with a typical phenotype [23]. The p.Arg71His (c.212G>A) variant was first described in two sisters with the mild form, compound heterozygous for p.Arg71His and p.Ala305Glu, who presented at 50 and 19 months with developmental delay, without macrocephaly, hypotonia, spasticity or seizures; the elder had mild cognitive and social impairment and the younger age-appropriate language and behaviour, and brain NAA was lower than expected in the classic form despite very deficient in vitro enzyme activity [6, 34]. Velinov and colleagues described a 28-month-old Ecuadorian girl, homozygous for p.Arg71His, with mild motor and speech delay at 9 months, symmetrical basal ganglia hyperintensities at 18 months, walking at 19 months, about 20 words at 25 months, without macrocephaly or seizures and with NAA about 15 times above normal [6]. Janson and colleagues noted that other protracted courses described did not appear to correlate with enzyme activity [6]. At the other extreme, p.Asp249Val was associated with manifestation at birth and early death [6].

In the German and North American cohort of Bley and colleagues, no formal genotype–phenotype correlation was demonstrated, but affected siblings had a similar course, suggesting the involvement of as yet unknown modifiers [10]. GeneReviews also records phenotypic concordance between affected relatives [1]. Deep mutational scanning of thousands of aspartoacylase variants showed a correlation between protein degradation and cellular toxicity, providing a resource for the functional interpretation of missense variants and suggesting that many act through protein instability [39, 40].

2.8. Allelic disorders and modifiers

No other diseases associated with ASPA variants are known [1]. No validated modifier genes have been described. The biochemical mirror image of the disease, hypoacetylaspartia due to NAT8L deficiency, with absent NAA, is a distinct entity caused by another gene (Section 6) [2].

3. Clinical Spectrum and System-Specific Manifestations

3.1. Sources of natural history data

The most complete description of the typical form comes from the cohort of Bley and colleagues, published in 2021: 23 patients followed prospectively and retrospectively, living mainly in Germany and then in the US, 20 of non-Jewish origin, compared with 59 patients from the literature and combined in a survival analysis of 82 patients [10]. The cohort is dominated by typical forms, and the authors themselves acknowledge that patients with milder phenotypes are less likely to reach their centre [10]. The natural history study CANinform (NCT04126005), sponsored by Aspa Therapeutics, included 67 participants in five age cohorts, with in-person or remote video motor assessments, and ended in August 2025, with no results published or posted on the registry by 26 October 2026 [41]. Another natural history study, with 21 participants in Boston, New York and Hamburg and gross motor function (GMFM) and magnetic resonance imaging (MRI) outcomes at 2 years, ended in 2019 [42]. The data that follow describe patients treated with supportive measures only, without disease-modifying therapy.

3.2. Typical form: onset and first year

Infants appear normal in the first weeks [1]. In the Bley cohort, symptoms appeared between 0 and 6 months (in the literature, median 3 months, range 0–18), and all patients had between 2 and 12 symptoms in the first 6 months, median 6; the most frequent were developmental delay (17/23), macrocephaly (12/23) and abnormal eye movements (12/23) [10]. Diagnosis was made between 0 and 29 months [10]. GeneReviews describes the triad of hypotonia, lack of head control and macrocephaly from 3–5 months, with poor visual tracking and nystagmus, sucking and swallowing difficulties and epileptic seizures [1]. Saudubray places onset mainly between 2 and 4 months and describes progression to marked delay, seizures, optic atrophy, progressive spasticity and opisthotonic posturing [2].

Tone evolves from hypotonia to spasticity [1]. In the Bley cohort, 8/23 patients had hypotonia only, 6/23 hypertonia only and 4/23 axial hypotonia combined with appendicular hypertonia at the time of assessment [10]. Spasticity may resemble cerebral palsy [2]. The OMIM clinical synopsis also records opisthotonus, generalised seizures, very early loss of milestones and late decerebrate or decorticate posturing, with a normal first month and onset between 2 and 4 months [6].

3.3. Macrocephaly

Head circumference is usually normal at birth [1, 2], and OMIM records late closure of the anterior fontanelle [6]. In the Bley cohort macrocephaly was present in all patients (91% in the literature, with median onset at 5 months, range 0–60), with onset between 0 and 19 months, median 7 months in girls and 8.5 months in boys, and a plateau at around 18 months [10]. GeneReviews describes progressive macrocephaly between 4 and 18 months in most patients [1], and Saudubray notes crossing of centiles after 6 months, with macrocephaly evident at one year [2]. Absence of macrocephaly does not exclude the diagnosis: among the nine patients with cytotoxic oedema in whom head circumference was described, one was microcephalic, five normocephalic and four macrocephalic [27].

3.4. Psychomotor development

Children with the typical form do not achieve sustained unsupported sitting, standing, walking or speech; more than half develop a social smile and laughter, and a minority reach for objects and use their hands purposefully [1]. The table summarises the milestones in the Bley cohort [10].

Three regularities have prognostic value. No child who was not sitting at 12 months went on to achieve that position; fine motor skills were retained longer than gross motor skills; and language comprehension persisted longer than expression [10]. All patients eventually lost the skills they had acquired [10].

3.5. Epilepsy

Seizures are of several types: tonic spasms, generalised tonic-clonic seizures and infantile epileptic spasms syndrome, described in series from Iran and Jordan [1, 43, 44]. In the Bley cohort, two thirds of patients had no seizures in the first year, the frequency increased significantly over time, peaking at the end of the first decade, and 57% of patients in the literature developed seizures [10]. Ten of 14 patients with seizures needed antiepileptic drugs and half of these needed polytherapy [10]. Seizures frequently persist despite treatment [1]. An acoustic startle was reported by more than half of the carers, and the authors consider that both tonic spasms and the startle response may be seen in the disease [10].

3.6. Vision and hearing

Early reduction in fixation and tracking is, in most cases, due to cerebral visual impairment, that is, bilateral impairment without an ocular cause and with normal ocular function [1]. In the Bley cohort only two patients had optic atrophy, and the authors attribute the visual impairment mainly to white matter disease of the visual pathways [10]. Saudubray and GeneReviews mention optic atrophy as a finding of the disease, and Saudubray includes the combination of optic atrophy and macrocephaly among the ocular signs pointing to this diagnosis [1, 2, 45]. Nystagmus appears early [1]. Hearing was preserved in all patients in the Bley cohort, unlike in the mouse model [10]. The OMIM clinical synopsis nonetheless includes deafness, based on the temporal bone study of two siblings born to consanguineous parents, which showed bilateral absence of the organ of Corti, supporting cells and hair cells, with mild secondary atrophy of spiral ganglion neurons and normal auditory nerves and vestibular organs (Ishiyama and colleagues, 2003) [3, 6].

3.7. Feeding, growth and respiratory complications

Chewing and swallowing deteriorate, with gastro-oesophageal reflux, vomiting and aspiration, leading to recurrent respiratory infections and poor weight gain [2]. In the Bley cohort, 10 of 22 patients (45%) had a gastrostomy, placed at a median age of 22 months (range 10–179), earlier than in the literature (52%, median 39 months, range 6–228), a difference the authors attribute to the evolution of palliative care [10]. With progression, irritability, sleep disturbance and increasing hypertonia appear [1, 2].

3.8. Atypical or mild form

The spectrum is broader than in the typical form. Speech and motor skills may be normal or mildly delayed at first, many children have a normal head circumference and most attend school with an individual education plan [1]. Ataxia and impaired coordination and gait are frequent [1, 34, 35, 46–48]. Progress is slow, without regression until late stages; many children achieve walking, and some have significant language impairment and a diagnosis of intellectual disability [1, 48, 49]. Pigmentary retinopathy and epileptic seizures [50], a stroke-like presentation [51] and severe retinal degeneration [52] have been described. In 2025 a 13-year-old girl was described with intention tremor and fine motor difficulties, whose symmetrical multiregional MRI initially suggested a neurometabolic or mitochondrial disorder, and whose diagnosis was confirmed by biochemistry and genome sequencing [11]. Two non-Jewish siblings with protracted N-acetylaspartic aciduria had identical clinical and MRI pictures [53].

Life expectancy in the mild form is poorly characterised, and patients may reach adulthood [1]. In the Telugu Devanga Chettiar community in southern India, about 24 cases of the mild or juvenile form associated with the founder variant c.526G>A have been recorded historically [35].

3.9. Congenital and late juvenile forms

Saudubray describes a congenital variant, with symptoms at birth or shortly after, and a juvenile variant with onset after 5 years [2]; Matalon and colleagues already referred in 1989 to congenital, infantile and late-onset forms [6]. The two British patients with p.Asp249Val manifested the disease in the neonatal period, with a severe course and early death [3, 6]. The sources consulted do not provide cohorts that characterise these variants separately.

3.10. Neuropathology

At autopsy, the white matter is soft and gelatinous. Microscopy shows progressive loss of myelinated arcuate fibres and vacuolation of the deep layers of the grey matter and of the subcortical white matter, with the central white matter relatively spared [2]. White matter vacuolation and demyelination produce the spongy appearance that gives the disease its name [1].

3.11. Survival: natural history and treated disease

Classic descriptions gave a survival of about 3 years, with rare exceptions up to 10 [10]. Saudubray, in the 2022 edition, still places death usually in the first decade, with survival possible into the second decade in a vegetative or near-vegetative state [2]. In the pooled analysis of Bley (82 patients, cohort and literature), 73% of patients reached 10 years; 5 of the 23 cohort patients and 21 of the 59 from the literature died, with no significant difference in survival between the two groups [10]. Neither epileptic seizures nor the acquisition of visual tracking or head control in the first 2 years significantly influenced survival [10]. GeneReviews concludes that most children with the typical form die in the first two decades and that survival is longer than previously described, probably owing to improved medical and nursing care [1]. Very long survival has been described from early on: a 33-year-old Ashkenazi Jewish woman, with intellectual disability and extrapyramidal cerebral palsy diagnosed at 2 years, blindness at 5 years with preserved pupillary reflex, generalised seizures from 8 years and bilateral optic atrophy at 20 years, was in a persistent vegetative state at 33.5 years; her younger sister, also affected, died at 9 years after aspiration (Feigelman and colleagues, 1991) [6]. The divergence between Saudubray and GeneReviews is resolved in favour of the 2021 data, which are more recent and quantified. These figures refer to patients receiving supportive care; data from the gene therapy trials do not yet allow effects on survival to be estimated (Section 7).

4. Clinical Recognition and Warning Signs

4.1. When to suspect

In the infant, suspicion should arise with the combination of axial hypotonia with lack of head control after 3 months, accelerated growth of head circumference crossing centiles, poor visual tracking or nystagmus, irritability and developmental stagnation [1, 10]. Bley and colleagues emphasise that macrocephaly associated with early developmental stagnation is the feature that distinguishes Canavan disease from many infantile encephalopathies, such as cerebral palsy, and from other neurodegenerative diseases, such as Krabbe disease [10]. The Oxford Desk Reference includes Canavan disease among the causes of macrocephaly with regression and among the leukodystrophies presenting with increased head circumference, alongside Alexander disease, megalencephalic leukoencephalopathy with subcortical cysts and the gangliosidoses [54, 55]. In the investigation of macrocephaly, defined as head circumference above 3 standard deviations, it recommends MRI, urinary organic acids and screening for mucopolysaccharidoses [56].

In the older child and adolescent, the mild form manifests as mild speech or motor delay, ataxia, intention tremor, fine motor difficulty, nystagmus or visual impairment, frequently with normal head circumference [1, 5, 11]. In these cases, MRI with hyperintensity of the basal ganglia, thalami and brainstem is easily interpreted as mitochondrial disease [1, 11].

4.2. Red flags and urgent settings

4.3. Newborn screening

Canavan disease is not on the US Recommended Uniform Screening Panel (RUSP), either among the core or the secondary conditions, in the January 2023 version published on the page revised in July 2024 [59]. Among other criteria, the RUSP requires the existence of effective treatment [59]. Technically, an LC-MS/MS method for NAA in dried blood spots separated patients and controls without overlap: in affected newborns the median was 7.3 µmol/L (range 5.1–9.9; n=8) and in neonatal controls 0.93 µmol/L (0.4–1.8; n=784) [60]. The authors consider the method suitable for newborn screening and multiplexable, but make population application conditional on the availability of a promising treatment [60]. The Early Check study of expanded newborn screening, coordinated by RTI International, planned 30,000 newborns, is active and not recruiting and has no Canavan-specific results posted [61]. The sources consulted do not describe newborn screening programmes for the disease in Europe.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria

No published consensus clinical diagnostic criteria exist [1]. The diagnosis is established in a proband with suggestive findings by demonstrating raised NAA in urine, by GC-MS, or in the brain, by proton MRS, and/or by identifying biallelic pathogenic or likely pathogenic variants in ASPA [1]. Urinary NAA elevations of the order of a hundred times normal are considered pathognomonic, but Saudubray recommends molecular and/or enzymatic confirmation [2]. The presence of biallelic variants of uncertain significance (VUS), or of one pathogenic variant and one VUS, neither establishes nor excludes the diagnosis, and urinary NAA or MRS help to interpret these variants [1].

5.2. Biochemistry

Urinary NAA is the preferred biochemical method [1]. In the typical form it is frequently raised more than 100-fold; in the atypical form the rise is smaller, of the order of 10-fold [1]. In the Nagy series, mean urinary NAA was 525.3 mmol/mol creatinine (range 25.2–1335) in the mild form and 1369 mmol/mol creatinine (391.7–2420) in the typical form (p<0.0001), showing overlapping ranges and a distribution that correlates with phenotype [23]. A mild rise in NAA with macrocephaly may pose a diagnostic problem [62]. NAA is also raised in blood and cerebrospinal fluid (CSF) in the typical form, but urinary NAA is sufficient for diagnosis of this form [1, 55]. In the urinary organic acid profile, isolated N-acetylaspartate points to Canavan disease [63]. NAA in dried blood spots discriminated patients (median 5.7 µmol/L; 1.6–13.6; 45 samples from 25 patients) from controls (0.44; 0.24–0.99; n=59) without overlap and may come to have a diagnostic role beyond screening [1, 60].

In the 1989 Matalon series, all 21 patients had urinary NAA about 200 times that of controls and obligate carriers, a patient CSF sample contained 232 µmol/L of NAA (undetectable in the control), and urinary NAA was not raised in other leukodystrophies, such as Alexander and Krabbe diseases, metachromatic leukodystrophy and adrenoleukodystrophy [6]. In obligate carriers, aspartoacylase activity was less than 50% of control values [6]. Aspartoacylase activity in cultured fibroblasts fluctuates with culture conditions and may be undetectable in both the typical and the atypical form, so it does not distinguish the forms and is not used routinely [1, 64].

5.3. Imaging and spectroscopy

Computed tomography (CT) and MRI in early infancy may be read as normal [1, 65]. The typical abnormality is a diffuse, symmetrical leukodystrophy of the subcortical white matter, extending to cortical areas, the brainstem and the cerebellum, with T1 hypointensity and T2 hyperintensity [1, 2, 66]. The globus pallidus is involved and the caudate and putamen are spared [1]. In an Iranian series of seven patients studied by exome sequencing, the deep and subcortical white matter and the globus pallidus were affected in all cases, and the putamen, caudate and claustrum were spared [67]. In the Dominguez Gonzalez series, diffuse, symmetrical white matter change was associated with variable areas of reduced diffusivity, with involvement of the globus pallidus, thalamus and dentate nuclei and relative sparing of the striatum, corpus callosum and small regions of the posterior limb of the internal capsule [12]. An image in Saudubray shows, in a 6.5-year-old girl, the discrepancy between the severely affected subcortical white matter and the relatively spared central white matter, at least in the frontal regions [2].

Cytotoxic oedema, with restricted diffusion, is frequent in the initial stages. Merrill and colleagues identified 13 patients with this finding in the literature, including their own case, with a median age at presentation of 12 months (mean 32; range 0.8–171), of whom 10 of 11 with recorded sex were male [27]. The structures most affected were the subcortical white matter, present in all, the brainstem, basal ganglia, thalamus, globus pallidus, cerebellum, internal capsule and corpus callosum [27]. As restricted diffusion is visible for about a week after injury, its presence indicates that part of the injury occurs in the postnatal period [27].

In the atypical form, MRI may not show diffuse changes, and subcortical involvement has been described [1, 46, 48, 53]. T2 and FLAIR (fluid-attenuated inversion recovery) hyperintensities are common in the basal ganglia, thalami, brainstem, especially the pontine tegmentum, and cerebellum [1, 47, 49, 68]. A juvenile form without white matter changes has been described [69].

Proton MRS is one of the few imaging investigations with diagnostic value in metabolic diseases, because it shows an increased NAA peak at 2.02 ppm [27, 45]. It has been described as diagnostic in infants even with normal serum and urinary NAA [1, 43]. In a retrospective accuracy study with 18 patients and 10 controls, at an echo time (TE) of 135 ms, the proposed cut-offs were 2.75 for the NAA/creatine (Cr) ratio and 7.18 for NAA/choline (Cho), with a corrected area under the curve of 0.96 (95% confidence interval, 95% CI: 0.95–1.00) after bootstrap validation; the authors present them as preliminary single-centre thresholds [70]. In Merrill's case, the NAA/Cr ratio at short TE was 2.45, against a normal value for age of 1.71, which shows that thresholds depend on TE and age [27].

5.4. Molecular tests and algorithm

GeneReviews proposes two approaches [1]. When the clinical and laboratory picture suggests the disease, single-gene testing begins with ASPA sequencing; if one or no variant is identified, targeted deletion and duplication analysis follows; with a biochemical diagnosis and one or no variant, comprehensive genomic testing for deep intronic or structural variants is considered [1]. Leukodystrophy or cerebral palsy panels that include ASPA are an alternative [1]. When the phenotype is non-specific, the exome is the most used test, and the genome is indicated for deep splice and structural variants when suspicion is high and the exome is negative [1]. Biochemistry and genetics are frequently requested in parallel, and the choice also depends on laboratory turnaround time [1].

The June 2026 GeneReviews revision adds that patients with no variants or only one variant after short-read exome or genome sequencing should be considered for testing for the SVA_E insertion [1]. This can be identified by long-read sequencing, with or without CRISPR/Cas9 enrichment, or by a targeted PCR that yields 260 bp from the normal allele and about 3000 bp from the allele with the insertion [9, 12]. RNA sequencing confirms the splicing effect [12].

5.5. Limitations of testing

The proportion of variants detectable by a method, such as the 99% attributed to sequencing in the GeneReviews table, is not the clinical sensitivity of that method [1]. In the Hamburg cohort, about 13% of patients lacked a complete molecular diagnosis before testing for the SVA_E insertion [9]. A negative molecular result does not exclude the disease in a patient with raised urinary NAA, and parental mosaicism and uniparental isodisomy should be considered when segregation is not as expected [1].

5.6. Report content and interpretation

The molecular report should identify the gene, the transcript with version (NM_000049.4), the Human Genome Variation Society (HGVS) nomenclature of each variant at DNA and protein level, zygosity and the classification according to ACMG/AMP criteria [4, 36]. For the SVA_E insertion, as for other structural variants, the detection method and the genomic coordinates with the reference genome version should be clear [9]. Confirmation of phase by parental testing is needed to distinguish compound heterozygosity from two variants in cis [1]. ASPA VUS can be interpreted with urinary NAA or MRS, which act as functional evidence in the patient, and with deep mutational scanning data [1, 39]. When the molecular diagnosis is not confirmed, DNA banking is recommended [1].

5.7. Prenatal and preimplantation diagnosis

When both familial variants have been identified, molecular prenatal diagnosis and preimplantation genetic testing are possible [1]. Aspartoacylase activity in amniocytes and chorionic villi was the historical basis of enzymatic prenatal diagnosis [17], but Kaul and colleagues noted in 1993 that this diagnosis is difficult, because activity is low or undetectable in normal direct or cultured chorionic villi and in normal cultured amniocytes [6]; prenatal diagnosis is now molecular. Counselling aspects are covered in Section 8.

6. Differential Diagnosis

6.1. Comparative table

In the table, AR denotes autosomal recessive inheritance and AD autosomal dominant inheritance.

6.2. Main differentials

The differential is organised around four elements: macrocephaly, subcortical leukodystrophy, the NAA peak and, in the mild form, involvement of the deep nuclei. Among diseases with macrocephaly and white matter change, Saudubray includes glutaric aciduria type 1, L-2-hydroxyglutaric aciduria, Canavan disease, the mucopolysaccharidoses, MLC, leukoencephalopathy with vanishing white matter and anterior-predominant Alexander disease [45]. Among small-molecule diseases with macrocephaly, Canavan disease is one of the three main ones, with glutaric aciduria type 1 and L-2-hydroxyglutaric aciduria [45]. The combination of macrocephaly with startle to sounds, incessant crying, myoclonus and irritability points to GM2 gangliosidosis, Canavan disease, Alexander disease and infantile Krabbe disease [45]. Family history and Ashkenazi Jewish ancestry strengthen suspicion, but most current diagnoses occur in people without that ancestry [1, 54].

Within the group of neurodegenerative diseases with macrocephaly, developmental delay, white matter involvement and visual changes, the distinction is made by urinary NAA measurement or MRS [10]. Alexander disease is distinguished by its frontal gradient, and MLC by ataxia, occasional seizures and mild cognitive decline [1]. The imaging pattern of Canavan disease, with involvement of the globus pallidus and sparing of the caudate and putamen, is useful in separating it from Leigh syndrome and the mitochondrial disorders with striatal lesions [1, 67].

Two pitfalls deserve emphasis. The first is restricted diffusion in an infant, which may lead to investigation for stroke; Merrill and colleagues point out that several patients with Canavan disease were initially investigated on that suspicion [27]. The second is the mild form, whose MRI with bilateral hyperintensity of the basal ganglia, thalami and pontine tegmentum leads to a mistaken diagnosis of mitochondrial disease [1, 11]. In both settings, urinary NAA measurement or MRS settles the question.

7. Treatments and Risk Management Strategies

7.1. General principles

There is no cure or approved therapy for Canavan disease, and no published clinical practice guidelines exist; the GeneReviews recommendations are based on the authors' experience [1, 8]. ERN-RND has not published specific recommendations for this disease, only general diagnostic flowcharts for leukodystrophies and documents for other entities [15]. Treatment is supportive, multidisciplinary and common to the typical and atypical forms, with three central aims: adequate nutrition and hydration, management of infection risk and airway protection [1]. No randomised trial has been carried out in the disease, and the available evidence is limited to expert opinion, natural history studies and small open-label trials [1, 10].

7.2. Supportive treatment by manifestation

Paediatric palliative care is part of the plan from the diagnosis of the typical form. In the Bley cohort, the earlier placement of gastrostomy, compared with the literature, was attributed to the improvement in such care, and survival longer than in older series was attributed to the overall improvement in medical and nursing care [1, 10]. The educational and social support interventions described in GeneReviews refer to the US system (early intervention from 0 to 3 years, preschool education from 3 to 5 years, individualised education programmes); the national equivalents apply in each country [1].

7.3. Symptomatic drugs: dosing and safety

There are no doses established specifically for Canavan disease. The following doses come from the official labels for spasticity and for epilepsy in general; their use in Canavan disease is an extrapolation from the general indication, without efficacy data in the disease [1].

Irritability and sleep disturbance, frequent in the typical form [1, 2], should be taken into account when choosing an antiepileptic drug whose adverse effects include behavioural changes, such as levetiracetam [74]. Levetiracetam is also the seizure prophylaxis drug used in Myrtelle's intracerebroventricular gene therapy protocol, at a dose of 20–50 mg/kg/day in two doses, orally or by gastrostomy, postoperatively and for 3 months [76].

7.4. Experimental metabolic therapies

Two pharmacological strategies have been studied on the basis of the pathophysiological hypotheses: acetate supplementation, to compensate for the lack of acetate for myelin synthesis, and reduction of brain NAA concentration. Neither has shown clinical benefit [2, 10].

Glyceryl triacetate (GTA) is an acetate precursor. In a low-dose trial, two infants aged 8 and 13 months received oral GTA for 4.5 and 6 months, starting at 25 mg/kg twice daily and doubling weekly up to a maximum of 250 mg/kg, without detectable toxicity and without clinical deterioration; in parallel, much higher doses in the tremor rat caused no serum biochemical changes or histopathological lesions [77]. In a subsequent high-dose trial, two infants aged 8 months and 1 year received 4.5 g/kg/day for 4.5 and 6 months, without significant adverse effects and without motor improvement; the authors attributed the lack of response to late initiation and advocated studies before 3 months of age [78]. Two Israeli GTA registrations, a phase 1 trial in two infants under 15 months and another study in a newborn, have unknown status, the first without an update since 2006 [79]. Dietary triheptanoin reduced oxidative stress, preserved oligodendrocytes, increased myelin and improved motor function in nur7 mice treated in the neonatal period, with much smaller effects when started later; there are no human studies in the sources consulted [80].

Lithium citrate was used to reduce brain NAA. In the first patient treated, the spectroscopic and clinical changes observed during administration reversed during a 2-week washout period after discontinuation [81]. In six patients treated for 60 days, lithium was safe, without clinical or laboratory toxicity; the parents reported improved alertness and social interaction, the GMFM did not improve to a statistically significant degree, MRS showed an overall fall in NAA, significant in the basal ganglia, and MRI suggested a slight improvement in frontal myelination [82]. The doses used in these two studies are not given in the abstracts consulted. A French phase 2 trial that planned lithium gluconate 1.3 mEq/kg/day in three doses combined with GTA 500 mg/kg/day in four doses, in 18 patients aged 1 to 15 years, was withdrawn without any participant enrolled, so these protocol doses were never administered in that study [83].

In the Bley cohort, 11 of 23 patients temporarily received medication without an approved indication, such as lithium, acetazolamide, calcium acetate, amoxicillin, sodium succinate and acetyl-L-carnitine [10]. Three reported improved alertness and mood stabilisation with lithium, but no patient had improvement in psychomotor abilities, neurological symptoms or head growth, and 5 of the 11 had adverse effects, mainly gastrointestinal or reduced alertness [10]. Saudubray concludes that lithium is safe but without clinical efficacy and that the efficacy of GTA has not been demonstrated [2]. None of these substances should be presented to families as an established treatment; their use outside a trial is off-label and without demonstrated benefit.

Other strategies remain preclinical: reduction of NAA synthesis by inhibitors of aspartate N-acetyltransferase (NAT8L), identified in a high-throughput screen [84], and pharmacological chaperones capable of selectively stabilising mutant aspartoacylases [85].

7.5. Gene therapy

7.5.1. First generation: intraparenchymal AAV2

The Leone and Janson programme began with the transfer of the aspartoacylase gene to the mammalian CNS and with a clinical protocol for neurosurgical administration of AAV2-ASPA [21, 22]. In the phase 1 study, the vector was administered through six burr holes, with a maximum dose of 1×10^12 vector genomes per patient; of 10 patients followed, 7 did not develop serum neutralising antibodies against AAV2 and 3 had low to moderately high titres, with minimal systemic signs of inflammation and CSF without neutralising antibodies or signs of brain inflammation [86]. In the long-term follow-up of a prospective cohort of 28 patients, 13 treated with 9×10^11 vector genomes at six intraparenchymal sites, there were no long-term adverse effects attributable to the vector over at least 5 years, and the authors described a fall in brain NAA, slowing of brain atrophy, some improvement in seizure frequency and stabilisation of overall clinical status, in an analysis by generalised linear mixed model of surrogate markers and clinical subscales [87]. The natural history by MRS and diffusion-weighted MRI of 28 patients, compared with paediatric reference values, provided the methodological basis for the imaging endpoints of these trials [88].

7.5.2. rAAV-Olig001-ASPA (MYR-101, Myrtelle)

rAAV-Olig001 is a recombinant vector with selective tropism for oligodendrocytes. In the open-label phase 1/2 trial NCT04833907, eight children with the typical form received an intracranial dose of 3.7×10^13 vector genomes of rAAV-Olig001-ASPA (MYR-101), with an interim analysis at 12 months, published in Nature Medicine in September 2025 [13]. All participants were of white ethnicity and 5 (62.5%) were male [13]. All had at least one adverse event and 7 (87.5%) had at least one serious adverse event, none considered related to MYR-101, and all resolved completely [13]. Treatment reduced CSF NAA (p=0.0008), increased myelination (p=0.0137) and improved developmental outcomes on the Mullen Scales of Early Learning (p=0.0171); the authors classify these results as preliminary [13]. Five of the seven authors declared financial ties to or roles at the sponsor [13].

The registry describes a single intracerebroventricular dose at two predefined sites, in sequential age cohorts (over 36 to 60 months; 15 to 36 months; under 15 months, with a minimum age of 3 months), up to 24 children, at a single centre in Dayton, Ohio [76]. It excludes children with previous intracranial surgery, previous gene therapy, absolute contraindication to immunosuppression or to MRI, vaccination in the previous month, life expectancy under 12 months or a GMFM-88 score above 35% [76]. The primary endpoint is safety at 12 months by the Common Terminology Criteria for Adverse Events (CTCAE) v5.0; the secondary endpoints include myelination by synthetic MRI, brain NAA by MRS compared with historical controls, GMFM-88, Mullen Scales, spasticity, seizures and CSF NAA at 6 months [76]. The protocol includes seizure prophylaxis with levetiracetam (20–50 mg/kg/day in two doses for 3 months) and prednisone tapered over 3 months to prevent late immune responses: 0.5 mg/kg/day in weeks 1–4 and 0.3 mg/kg/day in weeks 5–8, followed by a tapering dose in weeks 9–12 whose unit the registry does not specify, with extension of the taper by a further 3 months or switch to steroid-sparing immunosuppression if new T2 FLAIR inflammation appears at 3 months [76]. The registry status is enrolling by invitation only, with primary completion expected in August 2026 and study completion in August 2027 [76].

7.5.3. BBP-812 (Aspa Therapeutics, BridgeBio)

BBP-812 is a recombinant AAV9 vector carrying the ASPA transgene under a ubiquitous promoter, for expression in neuronal and non-neuronal cells, administered by a single intravenous infusion [89]. The CANaspire trial (NCT04998396) is an open-label, non-randomised phase 1/2 study with two dose levels followed by an expansion phase at the selected dose, and plans 26 participants [89]. It includes children up to 30 months with raised urinary NAA, biallelic ASPA variants and active clinical signs, and excludes patients with positive total anti-AAV9 antibodies, previous AAV therapy, high-dose immunosuppressants or very advanced disease, defined by continuous decerebrate or decorticate posturing, recurrent status epilepticus or seizures refractory to three or more antiepileptic drugs [89]. The primary endpoints are adverse events up to week 52 and the change in urinary NAA and brain NAA by MRS at 12 months; the secondary endpoints include GMFM-88, Bayley-4 and Vineland-3 [89]. The registry indicated active recruitment in April 2026, with primary completion expected in October 2026 [89].

In January 2026, in an investor presentation filed with the Securities and Exchange Commission (SEC), the sponsor stated that the trial is now pivotal, that BBP-812 reduces urinary NAA in a dose-dependent manner to levels associated with mild disease, that the FDA is open to the use of urinary NAA as a surrogate endpoint for accelerated approval and that trajectory analysis shows dose-dependent separation in the total score of the Hammersmith Infant Neurological Examination, part 2 (HINE-2), compared with the natural history study, with improvement in sitting, head control and reaching and grasping; the data cut-off was 10 December 2025, with 18 participants, and the announced path points to a biologics licence application (BLA) in 2027 [14]. These data have not been published in a peer-reviewed journal, have no numerical values or confidence intervals and constitute sponsor information [14]. The 2025 annual report describes the programme as designed for accelerated approval from a single registrational study with a biomarker, by the intravenous route and without neurosurgery [8]. The quarterly report of August 2026 adds no clinical data [90].

7.5.4. Other programmes, expanded access and discontinued trials

A single-patient expanded access protocol at the University of Florida planned rAAV9-CB6-AspA by simultaneous intravenous and intracerebroventricular routes, in a boy aged 18–24 months, with prior immunomodulation with rituximab and sirolimus; the registry has listed it as no longer available since 2023, without published results [91]. A phase 3 trial of ALD-101 as an adjunct to umbilical cord blood transplantation in inherited metabolic diseases, which included Canavan disease, was terminated by sponsor decision [92].

In preclinical research, neural and oligodendrocyte progenitor cells derived from induced pluripotent stem cells (iPSC), with functional ASPA introduced by lentivirus, CRISPR/Cas9 or TALEN, transplanted into Nur7 mice, reconstituted enzyme activity, reduced brain and CSF NAA and corrected the spongiform degeneration, myelination and motor function, with prolonged survival [93, 94]. In 2026, transplantation of these cells into already symptomatic mice, on postnatal day 21, reversed established pathology and improved motor function at 6 months, in contrast with earlier approaches, effective only in the presymptomatic phase [95].

7.6. Clinical trials: summary

The gene therapy trials are open-label, without a concurrent control group, and compare results with historical controls or with natural history studies, which makes the clinical endpoints sensitive to selection bias and to the heterogeneity of the natural history [14, 76]. The endpoints with statistical significance are mainly biomarkers (NAA in CSF, urine and brain) and imaging; the reduction in NAA demonstrates enzyme expression, but its relationship with long-term functional benefit is not yet established [13]. The two programmes differ in vector, cellular target, route of administration, and the age and severity of the patients included, so their results cannot be compared indirectly. The 2025 publication concerns eight patients, was read only as an abstract and does not provide confidence intervals; the BBP-812 data exist only in sponsor communications [13, 14]. The metabolic trials with GTA and lithium were all negative for clinical benefit, and one phase 3 trial and one phase 2 trial in the field were terminated or withdrawn [78, 82, 83, 92].

7.7. Regulatory status

Orphan drug designations and designations under accelerated programmes are not equivalent to marketing authorisation or to proof of efficacy. The sources consulted do not describe authorisation procedures under way at the EMA, INFARMED or the Brazilian Health Regulatory Agency (ANVISA) for Canavan disease therapies.

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

7.9. Agents to avoid, pregnancy, perioperative care and transition

The sources consulted do not identify drugs specifically contraindicated in Canavan disease. The precautions derive from the labels and from clinical features: botulinum toxin requires caution in patients with dysphagia or compromised respiratory function, given the risk of spread of the effect [58]; intrathecal baclofen cannot be stopped abruptly, and the risk is greater in patients with communication difficulties, as most patients with the typical form are [57]; oral baclofen and levetiracetam are always withdrawn gradually [72, 74]. The intravenous gene therapy trials exclude patients with previous AAV therapy, and the intracerebroventricular trial requires no vaccination in the preceding month, which should be taken into account when planning other interventions in families considering participation [76, 89].

Patients with the typical form do not reproduce; in the mild form no reproduction is known, but most patients described have not yet reached reproductive age [1]. There are no data on pregnancy or breastfeeding in affected women in the sources consulted.

The sources consulted contain no anaesthetic recommendations specific to the disease. The foreseeable perioperative risks arise from dysphagia, aspiration, respiratory failure and epilepsy [1]. Diagnostic and follow-up MRI frequently requires general anaesthesia or sedation, as in the case published by Boitsios and colleagues, and the authors of the Bley cohort considered sedation unjustifiable for research MRI outside an interventional context [10, 98].

Patients with the mild form may reach adulthood, and GeneReviews contains no specific guidance on transition to adult care [1]. Transition should ensure continuity of neurology, rehabilitation, nutrition and genetic counselling.

7.10. Guidelines and institutional positions

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

8.2. Carrier testing and cascade testing

The parents of a proband are presumed heterozygous, and confirmation in both is needed to estimate risk reliably and to confirm the phase of the variants [1]. Molecular carrier testing in relatives requires the familial variants to have been identified; biochemical detection is not used, because the fibroblast enzyme assay is complex and fluctuates with culture [1]. When one of the familial variants is the SVA_E insertion, the laboratory must use a method capable of detecting it, such as the targeted PCR described by Weiss and colleagues or long-read sequencing [9, 12].

The partner of a known carrier should be studied by full ASPA sequencing, because analysis targeted to the common variants is insufficient [1]. The Ashkenazi founder variants p.Glu285Ala, p.Tyr231Ter and c.433-2A>G account for 98.8% of mutant chromosomes in Ashkenazi Jews [19], but the proportion of alleles explained by targeted variants is far lower in other populations, in which the variants are diverse and private [2]. The SVA_E insertion, present in all gnomAD ancestry groups analysed, should be included in all carrier screening tests that cover ASPA, whether pan-ethnic or directed at a population [12].

8.3. Population carrier screening

The 2021 ACMG recommendation places ASPA in tier 3 of carrier screening, which groups genes with a carrier frequency of at least 1/200 in some population; the maximum frequency recorded for ASPA is 0.019856, about 1/50 [99]. ACMG recommends offering tier 3 screening to all pregnant people and those planning a pregnancy, considering tier 4 only in consanguinity of second cousins or closer or in the presence of a personal or family history, and not offering tiers 1 and 2 alone, because they do not assess all groups equitably [99]. The document argues for ethnicity-neutral screening, because self-reported ancestry is an unreliable basis and restriction by socially defined ethnic groups is inequitable and scientifically flawed [99]. Reports should state the ACMG tier, the panel content, the technical approach and the types of variant detectable, should not report residual risk and should routinely report only pathogenic and likely pathogenic variants; VUS are reported only in partners of identified carriers and with consent [99]. This last requirement has a direct consequence for ASPA, because a panel that does not state that it detects the SVA_E insertion has a blind spot for the most frequent variant [12].

ACOG, in Committee Opinion 691 of 2017, reaffirmed in 2025, recommends that information about carrier screening be given to all pregnant people and that screening and counselling preferably take place before pregnancy, so that the couple knows the reproductive risk and has all options available, including deciding whether to conceive, preimplantation genetic diagnosis and donor gametes [100]. If a person is a carrier, testing should be offered to their reproductive partner, and simultaneous screening of both is suggested when time to decide about prenatal diagnosis is short; if both are carriers, genetic counselling should be offered and prenatal diagnosis and assisted reproduction techniques discussed [100]. The carrier should be encouraged to inform relatives of the risk and of the availability of testing, but the clinician does not disclose the result without the patient's authorisation [100]. The family history of the patient and, where possible, of the partner includes the ancestry of relatives and consanguinity, defined as a union between second cousins or closer relatives; a person with a family history of a genetic condition should be offered screening for that condition [100]. Carrier screening for a condition is usually performed once in a lifetime, with the result recorded in the clinical record; as more recent panels include additional variants, the decision to repeat screening rests with a genetics professional, who assesses the incremental gain [100]. For Canavan disease, this rule has a practical consequence: screening performed before 2025 that did not detect the SVA_E insertion leaves a negative result of incomplete value, and it is in that context that targeted repeat testing may be discussed [12, 100]. Prenatal carrier screening does not replace newborn screening, nor does newborn screening replace the value of carrier screening [100]; in the case of Canavan disease, which is not part of the recommended uniform newborn screening panel in the US (RUSP), only carrier screening allows risk to be identified in advance [59]. If a person requests screening for a condition for which a test is available and which would be considered in another screening strategy, the test should be offered after counselling, regardless of ancestry and family history, and the cost of each option to the patient and to the health system should weigh in the choice, bearing in mind that a single test may cost more than a commercial expanded panel [100].

In the section devoted to people of Central and Eastern European Jewish ancestry, ACOG defines as Ashkenazi those people whose Jewish relatives originate from those regions and notes that most people of Jewish ancestry in North America descend from Ashkenazi communities [100]. ACOG had recommended offering carrier screening for four conditions in this population, Canavan disease, cystic fibrosis, familial dysautonomia and Tay-Sachs disease, and retains them in the 2017 document [100]. The clinical description accompanying the recommendation presents Canavan disease as a severe degenerative neurological disorder, of variable phenotype, with motor delay in the first months (head control and sitting), macrocephaly, hypotonia and intellectual disability, variable life expectancy with death frequently in childhood or adolescence and supportive treatment, caused by variants in the aspartoacylase gene, which takes part in the metabolism of N-acetyl-L-aspartate [100]. When only one member of the couple is Ashkenazi, that person should be screened first, and the other only if the first is a carrier; the couple should be informed that the carrier frequency and detection rate in non-Jewish people are unknown for most of these conditions, with the exception of Tay-Sachs disease and cystic fibrosis, which makes the couple's risk difficult to estimate [100]. For the same reason, counselling about residual risk after a negative screen in a non-Jewish person is complex, and ACOG suggests genetic counselling to define the best approach when one member of the couple is a carrier and the other has no Jewish ancestry [100]. The document also lists a broader panel for the Ashkenazi population, with less frequent conditions (carrier frequencies from 1/15 to 1/168), limited to conditions that meet the criteria of Committee Opinion 690 [100].

The two North American positions diverge in the basis for screening. ACOG in 2017 retains, for Canavan disease, the ancestry-based model, with testing offered on request outside that group [100]; ACMG in 2021 includes ASPA in tier 3, offered to all pregnant people without an ethnic criterion [99]. The divergence over the partner of a non-Jewish carrier is also one of method: ACOG refers to genetic counselling in the face of an unknown detection rate [100], whereas GeneReviews recommends full ASPA sequencing in that partner [1], which is the laboratory answer to the problem identified by ACOG. The 2008 ACMG technical standard for reproductive screening in the Ashkenazi Jewish population, which included Canavan disease, corresponds to the earlier, ancestry-based model [1, 20].

8.4. Prenatal diagnosis and preimplantation genetic testing

With both familial variants known, molecular prenatal diagnosis and preimplantation genetic testing are possible [1]. The ideal time to define risk and discuss options is before pregnancy [1, 100]. Views on prenatal diagnosis differ between professionals and families, and the decision is personal [1]. When the proband's molecular diagnosis is not confirmed, DNA banking preserves the possibility of future testing, including for structural variants that have only now become detectable [1, 12].

8.5. Counselling content and minors

Counselling should explain autosomal recessive inheritance, the similarity of phenotype between siblings, the distinction between the typical and the mild form and the absence of approved therapy, with information about gene therapy trials that require early diagnosis (up to 30 months in CANaspire) [1, 89]. GeneReviews recommends offering counselling to young adults who are carriers or at risk of being carriers [1]. For relatives at risk, GeneReviews refers to genetic counselling and does not propose assessment of asymptomatic siblings for early treatment [1]. The sources consulted contain no specific guidance on carrier testing in minors. Carriers should be informed of the risk to their relatives and of the availability of testing [100].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and what each figure measures

The allele frequency of the SVA_E insertion and the maximum credible allele frequency of 0.0013, calculated by the authors for an assumed prevalence of 1/10,000, are not prevalence estimates; the prevalence of 1/10,000 is a conservative assumption of the calculation [12]. The combination of a frequent pan-ancestral variant with other private variants suggests that the disease is underdiagnosed outside founder populations, and Aaron and colleagues allow that global prevalence may be underestimated because of the scarcity of community studies [7, 12]. The sources consulted do not provide a validated population estimate of prevalence in any European country or in the US.

9.2. Founder populations and geography

In the Ashkenazi Jewish population, p.Glu285Ala accounts for 82.9% to 84% of mutant alleles and p.Tyr231Ter for 13% to 14.8%, and these two variants, with c.433-2A>G, explain about 98.8% of mutant chromosomes [2, 19, 103]. The Oxford Desk Reference gives 80–85% for p.Glu285Ala and a carrier frequency greater than 1/40 [54, 56]. The frequency of these variants made targeted carrier screening possible [2]. Among 879 healthy Israeli Ashkenazi Jews, 15 were heterozygous for p.Glu285Ala (1/59) (Elpeleg, 1994), and among 1423 Ashkenazi Jews in Toronto screened for p.Glu285Ala, p.Tyr231Ter and p.Ala305Glu there were 25 carriers (1/57) (Feigenbaum, 2004) [3, 6]. In all Toronto carriers of p.Glu285Ala, the variant was in disequilibrium with the T allele of the polymorphism c.693C>T, indicating a founder chromosome [6]. The Jewish cases studied by Matalon traced back to a particular area of Eastern Europe [6].

In non-Jewish Europeans, p.Ala305Glu is the most frequently reported variant, with 60% of mutant chromosomes in the Kaul series, and the molecular basis in this population was characterised by Shaag and colleagues and by Elpeleg and Shaag [1, 19, 33, 104]. The SVA_E insertion was described in families of European, European-Uruguayan and European-American ancestry and confirmed in a German cohort [9, 12]. Most new diagnoses now occur in people without known Ashkenazi ancestry, probably because of the effect of widespread carrier screening in that population and the relative size of the populations [1, 10]. In the Bley cohort, 20 of the 23 patients were of non-Jewish origin, most often resident in Germany [10], and the Hamburg cohort of the PeriNAA study includes more than 55 patients [9].

In India, the Telugu Devanga Chettiar community has the largest known cluster of juvenile Canavan disease, with about 24 historical cases, genomic signatures of a founder effect without consanguinity (high fixation index, long runs of homozygosity and of identity by descent, long haplotype) and dating of the founder to about 375 years ago [35]. The carrier frequency of 1/23, confirmed by microsatellite markers, is higher than that described in the Ashkenazi Jewish population [7]. Outside these groups, OMIM records other clusters: about 35 cases identified in Saudi Arabia, where the disease appears unusually frequent (12 patients with enzyme deficiency in fibroblasts in the Ozand series, 1990), an Iranian family of first-cousin parents with four affected children out of nine (Mahloudji, 1970), deletion of exon 4 as a founder variant in Turkish patients and the variant p.Cys218Ter in homozygosity in Roma patients [3, 6]. In Iran, a series of seven patients studied by exome sequencing showed pathogenic or likely pathogenic ASPA variants in most [67], and the series from Iran and Jordan contributed to the description of the epilepsy [43, 44].

The sources consulted contain no prevalence data, carrier frequencies or patient series from Portugal or Brazil. The absence of data in these sources does not demonstrate the absence of published data.

9.3. Special populations

The mild form may reach adulthood, but life expectancy is poorly characterised and there are no data on pregnancy in affected women [1]. Children with the typical form are, in practice, paediatric patients with palliative care needs, and their carers need social support and respite [1]. In the gene therapy trials, eligibility depends on age and disease stage, which gives early diagnosis particular importance [76, 89]. All participants in the published MYR-101 trial were of white ethnicity, which limits the generalisability of the results [13].

9.4. Reference networks and patient associations

In Europe, Canavan disease falls within the scope of the leukodystrophies group of the European Reference Network for Rare Neurological Diseases (ERN-RND), although the group's page does not mention it by name, and within the scope of MetabERN, the network for inherited metabolic diseases [15, 16]. The European Leukodystrophy Association (ELA) is the European patient association listed in the trial registrations [76, 89]. In the US, the organisations referred to by GeneReviews and by the trial registries include the Canavan Foundation, the Cure Canavan Fund, the National Tay-Sachs and Allied Diseases Association (NTSAD), the Norton and Elaine Sarnoff Center for Jewish Genetics, the United Leukodystrophy Foundation and the Myelin Disorders Bioregistry Project [1, 89]. Research centres with natural history cohorts include the Massachusetts General Hospital, in Boston, and the University Medical Centre Hamburg-Eppendorf [1, 9, 42].

10. Illustrative Clinical Case

Merrill and colleagues described in 2016 a 10-week-old boy, born at term after an uncomplicated pregnancy, the son of non-consanguineous Caucasian parents, referred for episodes of right horizontal pendular nystagmus lasting 2 to 3 seconds [27]. Vital signs and general examination were normal, and head circumference was on the 63rd centile [27]. Neurological examination showed absent visual tracking, central hypotonia, limb hyperreflexia and choreiform movements [27].

Brain MRI at presentation showed extensive, symmetrical restricted diffusion in the subcortical white matter of both hemispheres, extending inferiorly along the corticospinal tracts to the brainstem, including the dorsal and ventral pons, with T2 hyperintensity and T1 hypointensity of the white matter [27]. Single-voxel MRS in the left posterior white matter showed an increased peak at 2.02 ppm, corresponding to NAA, with an NAA/Cr ratio at short TE of 2.45, against a normal value for age of 1.71 [27]. Urinary organic acids showed NAA excretion above 750 mmol/mol creatinine, and ASPA sequencing identified the variant c.914C>A, p.(Ala305Glu), in homozygosity, which confirmed the diagnosis [27]. Figure 1 of the article shows, in axial sections, the diffusion hyperintensity in the internal capsule and subcortical white matter with corresponding hypointensity on the apparent diffusion coefficient map, the T2 FLAIR, T2 and T1 images, and the MRS spectrum with the NAA peak marked (Merrill et al. 2016, Fig. 1; the description is based on the legend consulted) [27]. The article does not describe the subsequent course or management beyond diagnosis.

The case illustrates four points. The first is that presentation may precede macrocephaly: head circumference was normal at 10 weeks, and the review by the same authors found normocephaly in five of nine patients with cytotoxic oedema [27]. The second is that restricted diffusion, which lasts about a week after injury, indicates an active injurious process and supports the existence of a postnatal therapeutic window, an argument invoked by the gene therapy trials that include infants [27, 76, 89]. The third is that a pattern of restricted diffusion in an infant may be mistaken for stroke, and the NAA peak on MRS settles the diagnosis [27]. The fourth is that the p.Ala305Glu variant in homozygosity, although present in less severely affected patients in other series, may be associated with very early onset, which is consistent with the complete loss of activity in vitro described by Kaul and colleagues [10, 19].

A teaching case published in 2025 describes a 10-month-old girl, the second child of non-consanguineous parents, born at 41 weeks, with progressive macrocephaly, axial hypotonia, developmental delay and limb stiffness with brisk tendon reflexes suggestive of spastic diplegia; she did not hold her head at 4 months or sit at 10 months, her vocalisations expressed contentment or anger without age-appropriate babbling, eye contact was poor, she had no dysphagia or seizures, and prolonged electroencephalographic monitoring was normal; she underwent 3 T MRI with spectroscopy under general anaesthesia [98]. Only the clinical history of this case was consulted.

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

Life expectancy is reduced in all forms [1]. In the pooled analysis of 82 patients by Bley and colleagues, 73% reached 10 years, with no significant influence of seizures or of early acquisition of visual tracking or head control on survival [10]. Most children with the typical form die in the first two decades, and the sponsor of BBP-812 states that more than 25% of patients die by 10 years, a figure consistent with Bley's [1, 8]. Survival is longer than in the older series, which limited it to about 3 years, and that improvement is attributed to supportive care [1, 10]. In the typical form, absence of sitting at 12 months predicts that it will not be achieved, and all patients lose the skills they have acquired [10]. The mild form progresses slowly without regression until late stages, and patients may reach adulthood, but life expectancy has not been quantified [1]. Lower urinary NAA is associated with the mild phenotype and acts as a marker of severity [23].

11.2. Assessment instruments

The Canavan Disease Rating Score (CDRS) has 11 items (seizures; truncal hypotonia; appendicular spasticity; macrocephaly; feeding; language; responsive social smile; visual tracking; head control; reaching for objects; sitting without support), scored from 0 (absent or normal) to 2 (constant or marked), with a total of 0 to 22 [10]. The rules include, for example, 2 points for seizures with 3 or more seizures per year or the need for more than one antiepileptic drug, 1 point for macrocephaly when the circumference crosses 2 or more major centiles within the normal range and 2 points when it exceeds the 97th centile, and 2 points for feeding when tube feeding is required; interrater agreement in a pilot study was 96% (74 of 77 points) [10]. The trials also use the GMFM-88, part 2 of the HINE, the Alberta Infant Motor Scale, the Infant Motor Profile, the Mullen Scales of Early Learning, the Bayley-4 and the Vineland-3 [76, 89, 105]. Remote motor assessment by video was feasible in 95% (61/64) of assessments, with 86% acceptability to carers and good to excellent interrater reliability in 15 of 20 assessments [105], and agreement between remote and in-person assessment in 85 paired visits was almost perfect (Lin's concordance coefficient of 0.98 for the HINE-2 and for the GMFM-88) [97].

11.3. Initial evaluation after diagnosis

11.4. Surveillance

Follow-up MRI and MRS have no defined interval in GeneReviews outside a research context, and the need for sedation weighs in the decision [1, 10]. In the gene therapy trials, follow-up includes MRI, MRS, urinary and CSF NAA, and motor and developmental scales at 6 and 12 months [76, 89].

12. Summary and Concluding Remarks

Canavan disease is an autosomal recessive spongiform leukodystrophy caused by biallelic ASPA variants, which encodes the aspartoacylase of oligodendrocytes. Failure to hydrolyse the NAA produced by neurons causes NAA to accumulate in the brain, urine, blood and CSF and is associated with a lack of acetate for myelin synthesis, osmotic dysregulation and vacuolation of the subcortical white matter. The spectrum extends from the typical infantile form, with 85–90% of patients, to the mild or juvenile form, and severity correlates with residual enzyme activity and with urinary NAA.

The typical form begins in the first 6 months with axial hypotonia, lack of head control, progressive macrocephaly that plateaus at around 18 months, cerebral visual impairment with preserved hearing, spasticity, dysphagia and epilepsy that increases over the first decade. Few children sit, none walks without support, and all lose the skills they have acquired; with current care, 73% reach 10 years. The mild form manifests as mild delay, ataxia, tremor or visual impairment, often with normal head circumference and MRI of the deep nuclei that leads to a mistaken diagnosis of mitochondrial disease.

Diagnosis rests on urinary NAA by GC-MS or on the NAA peak on MRS and on the identification of biallelic ASPA variants. The discovery, in 2025, of an SVA_E insertion in intron 4, invisible to short-read sequencing and more frequent in population databases than any other pathogenic variant, makes it necessary to test for it by targeted PCR or long reads whenever sequencing finds one or no allele, and to include it in carrier screening panels. The founder variants p.Glu285Ala and p.Tyr231Ter explain most Ashkenazi alleles, p.Ala305Glu predominates in non-Jewish Europeans and c.526G>A causes a cluster of the juvenile form in a community in southern India; most current diagnoses occur outside the Ashkenazi population. ACMG recommends ASPA carrier screening for all pregnant people and those planning a pregnancy, without restriction by ethnicity.

There is no approved therapy and no clinical practice guideline. Treatment is supportive, with nutrition and airway protection, gastrostomy when swallowing fails, antiepileptic drugs, botulinum toxin and baclofen for spasticity, rehabilitation, augmentative communication and palliative care. Trials with glyceryl triacetate and lithium showed no clinical benefit. Gene therapy with vectors directed at oligodendrocytes by the intracranial route showed, in eight children, a reduction in CSF NAA and increased myelination with acceptable safety at 12 months, and the intravenous AAV9 vector BBP-812 is in a pivotal trial, with a dose-dependent reduction in urinary NAA according to the sponsor and a licence application announced for 2027. Eligibility for these trials depends on age and disease stage, which gives early recognition of the combination of hypotonia, macrocephaly and leukodystrophy, and urinary NAA testing, a clinical weight they did not have before.

13. References

1. Nagy A, Bley AE, Eichler F. Canavan Disease. 1999 Sep 16 [updated 2023 Dec 21; revised 2026 Jun 23]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID 20301412. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1234/
2. Kölker S, Hoffmann GF. Cerebral Organic Acid Disorders and Other Disorders of Lysine Catabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 22, sections 22.1, 22.7 (Fig. 22.4), 22.9 (Fig. 22.5), 22.10 and 22.12. doi:10.1007/978-3-662-63123-2_22
3. Online Mendelian Inheritance in Man, OMIM®. *608034 Aspartoacylase; ASPA [Internet]. Baltimore (MD): Johns Hopkins University; updated 25 Apr 2023, accessed 26 Oct 2026. Available from: https://www.omim.org/entry/608034
4. VariantValidator (version 4.0.1; vvdb_2025_3 database). Validation of ASPA variants on GRCh38 against NM_000049.4/NP_000040.1 [Internet]. Accessed 26 Oct 2026. Available from: https://rest.variantvalidator.org/
5. Orphanet. Canavan disease (ORPHA:141); Severe Canavan disease (ORPHA:314911); Mild Canavan disease (ORPHA:314918) [Internet]. Paris: INSERM; updated 2 Jul 2026, accessed 26 Oct 2026. Available from: https://www.orpha.net/en/disease/detail/141
6. Online Mendelian Inheritance in Man, OMIM®. #271900 Canavan disease (including Clinical Synopsis) [Internet]. Baltimore (MD): Johns Hopkins University; last edited 16 Jan 2025, accessed 26 Oct 2026. Available from: https://www.omim.org/entry/271900
7. Aaron R, Chapla A, Jayaprabha M, Elangovan J, Chellappa RK, Krishnamurthy A, et al. Prevalence and carrier frequency of Canavan disease in a South Indian community with implications for research and public health. Sci Rep. 2026;16(1):22983. doi:10.1038/s41598-026-52947-0
8. BridgeBio Pharma, Inc. Annual report on Form 10-K for the fiscal year ended December 31, 2025. Washington (DC): U.S. Securities and Exchange Commission; filed 24 Feb 2026. Accession no. 0001743881-26-000009. Available from: https://www.sec.gov/Archives/edgar/data/1743881/000174388126000009/bbio-20251231.htm
9. Weiß M, Selig M, Friedrich J, Wierczeiko A, Diederich S, Sigel H, et al. Deep intronic SVA_E retrotransposition as a novel factor in Canavan disease pathogenesis. Hum Gene Ther. 2025;36(19-20):1248-56. doi:10.1089/hum.2025.006
10. Bley A, Denecke J, Kohlschütter A, Schön G, Hischke S, Guder P, et al. The natural history of Canavan disease: 23 new cases and comparison with patients from literature. Orphanet J Rare Dis. 2021;16(1):227. doi:10.1186/s13023-020-01659-3
11. Rehsi P, Siddiqui A, Singh R, Ihezue O, Halligan R. A mild juvenile onset Canavan disease with atypical clinical presentation and MRI brain features. JIMD Rep. 2025;66(4):e70031. doi:10.1002/jmd2.70031
12. Dominguez Gonzalez CA, Bell KM, Rajagopalan R, de Silva MG, Lemes A, Zabala C, et al. Deep intronic SVA_E insertion identified as the most common pathogenic variant associated with Canavan disease: a diagnostic blind spot. Neurol Genet. 2025;11(5):e200291. doi:10.1212/NXG.0000000000200291
13. Leone P, Lober RM, Francis J, Flamini O, Cecil KM, Shera D, Janson CG. Oligodendrocyte-targeted adeno-associated virus gene therapy for Canavan disease in children: a phase 1/2 trial. Nat Med. 2025;31(11):3772-9. doi:10.1038/s41591-025-03919-w
14. BridgeBio Pharma, Inc. Current report on Form 8-K, Exhibit 99.2: corporate presentation, 44th Annual J.P. Morgan Healthcare Conference. Washington (DC): U.S. Securities and Exchange Commission; filed 12 Jan 2026. Accession no. 0001140361-26-000971. Available from: https://www.sec.gov/Archives/edgar/data/1743881/000114036126000971/ef20062706_ex99-2.htm
15. European Reference Network for Rare Neurological Diseases (ERN-RND). Disease knowledge hub: Leukodystrophies [Internet]. Accessed 26 Oct 2026. Available from: https://www.ern-rnd.eu/disease-knowledge-hub/leukodystrophies/
16. MetabERN. European Reference Network for Hereditary Metabolic Disorders [Internet]. Accessed 26 Oct 2026. Available from: https://metab.ern-net.eu/
17. Matalon R, Michals K, Sebesta D, Deanching M, Gashkoff P, Casanova J. Aspartoacylase deficiency and N-acetylaspartic aciduria in patients with Canavan disease. Am J Med Genet. 1988;29(2):463-71. doi:10.1002/ajmg.1320290234
18. Kaul R, Gao GP, Balamurugan K, Matalon R. Cloning of the human aspartoacylase cDNA and a common missense mutation in Canavan disease. Nat Genet. 1993;5(2):118-23. doi:10.1038/ng1093-118
19. Kaul R, Gao GP, Aloya M, Balamurugan K, Petrosky A, Michals K, Matalon R. Canavan disease: mutations among Jewish and non-Jewish patients. Am J Hum Genet. 1994;55(1):34-41. PMID 8023850. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC1918221/
20. Monaghan KG, Feldman GL, Palomaki GE, Spector EB; Ashkenazi Jewish Reproductive Screening Working Group. Technical standards and guidelines for reproductive screening in the Ashkenazi Jewish population. Genet Med. 2008;10:57-72. doi:10.1097/GIM.0b013e31815f6eac
21. Leone P, Janson CG, Bilaniuk L, Wang Z, Sorgi F, Huang L, et al. Aspartoacylase gene transfer to the mammalian central nervous system with therapeutic implications for Canavan disease. Ann Neurol. 2000;48(1):27-38. PMID 10894213. Available from: https://pubmed.ncbi.nlm.nih.gov/10894213/
22. Janson C, McPhee S, Bilaniuk L, Haselgrove J, Testaiuti M, Freese A, et al. Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum Gene Ther. 2002;13(11):1391-412. doi:10.1089/104303402760128612
23. Nagy A, Eichler F, Bley A, Bredow J, Fay A, Townsend EL, et al. Urine N-acetylaspartate distinguishes phenotypes in Canavan disease. Hum Gene Ther. 2024;36(1-2):45-56. doi:10.1089/hum.2024.168
24. Mendes MI, Smith DE, Pop A, Lennertz P, Fernandez Ojeda MR, Kanhai WA, et al. Clinically distinct phenotypes of Canavan disease correlate with residual aspartoacylase enzyme activity. Hum Mutat. 2017;38(5):524-31. doi:10.1002/humu.23181
25. Brivet M, Gaignard P, Schiff M. Disorders of Pyruvate Metabolism and the Tricarboxylic Acid Cycle. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Capítulo 11, secção 11.11 (Malate-Aspartate Shuttle defects). doi:10.1007/978-3-662-63123-2_11
26. Madhavarao CN, Arun P, Moffett JR, et al. Defective N-acetylaspartate catabolism reduces brain acetate levels and myelin lipid synthesis in Canavan's disease. Proc Natl Acad Sci U S A. 2005;102:5221-6. doi:10.1073/pnas.0409184102
27. Merrill ST, Nelson GR, Longo N, Bonkowsky JL. Cytotoxic edema and diffusion restriction as an early pathoradiologic marker in Canavan disease: case report and review of the literature. Orphanet J Rare Dis. 2016;11(1):169. doi:10.1186/s13023-016-0549-1
28. Gessler DJ, Li D, Xu H, Su Q, Sanmiguel J, Tuncer S, et al. Redirecting N-acetylaspartate metabolism in the central nervous system normalizes myelination and rescues Canavan disease. JCI Insight. 2017;2(3):e90807. doi:10.1172/jci.insight.90807
29. Becker I, Wang-Eckhardt L, Eckhardt M. NAAG synthetase deficiency has only low influence on pathogenesis in a Canavan disease mouse model. J Inherit Metab Dis. 2024;47(2):230-43. doi:10.1002/jimd.12693
30. Hull VL, Wang Y, McDonough J, Zhu M, Burns T, Al Ramel N, et al. Astroglial conditional Slc13a3 knockout is therapeutic in murine Canavan leukodystrophy. Ann Clin Transl Neurol. 2024;11(4):1059-62. doi:10.1002/acn3.52010
31. Cozzolino M, Augello B, Carella M, Palumbo O, Tavazzi B, Amorini AM, et al. Chromosomal 17p13.3 microdeletion unmasking recessive Canavan disease mutation. Mol Genet Metab. 2011;104:706-7. doi:10.1016/j.ymgme.2011.09.031
32. Stenson PD, Mort M, Ball EV, Chapman M, Evans K, Azevedo L, et al. The Human Gene Mutation Database (HGMD®): optimizing its use in a clinical diagnostic or research setting. Hum Genet. 2020;139:1197-207. doi:10.1007/s00439-020-02199-3
33. Shaag A, Anikster Y, Christensen E, Glustein JZ, Fois A, Michelakakis H, et al. The molecular basis of Canavan (aspartoacylase deficiency) disease in European non-Jewish patients. Am J Hum Genet. 1995;57:572-80. PMID 7668285. Available from: https://pubmed.ncbi.nlm.nih.gov/7668285/
34. Janson CG, Kolodny EH, Zeng BJ, Raghavan S, Pastores G, Torres P, et al. Mild-onset presentation of Canavan's disease associated with novel G212A point mutation in aspartoacylase gene. Ann Neurol. 2006;59:428-31. doi:10.1002/ana.20787
35. Kotambail A, Selvam P, Muthusamy K, Thomas M, Sudhakar SV, Ghati C, et al. Clustering of juvenile Canavan disease in an Indian community due to population bottleneck and isolation: genomic signatures of a founder event. Eur J Hum Genet. 2023;31(1):73-80. doi:10.1038/s41431-022-01198-4
36. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405-24. doi:10.1038/gim.2015.30
37. Zeng BJ, Wang ZH, Torres PA, Pastores GM, Leone P, Raghavan SS, et al. Rapid detection of three large novel deletion of the aspartoacylase gene in non-Jewish patients with Canavan disease. Mol Genet Metab. 2006;89:156-63. doi:10.1016/j.ymgme.2006.05.014
38. Caliebe A, Vater I, Plendl H, Gesk S, Siebert R, Cremer FW, et al. A 439 kb-sized homozygous deletion in 17p13.3 leading to biallelic loss of the ASPA as cause of Canavan disease detected by SNP-array analysis. Mol Genet Metab. 2010;99:184-5. doi:10.1016/j.ymgme.2009.10.011
39. Grønbæk-Thygesen M, Voutsinos V, Johansson KE, Schulze TK, Cagiada M, et al. Deep mutational scanning reveals a correlation between degradation and toxicity of thousands of aspartoacylase variants. Nat Commun. 2024;15(1):4026. doi:10.1038/s41467-024-48481-0
40. Grønbæk-Thygesen M, Hartmann-Petersen R. Cellular and molecular mechanisms of aspartoacylase and its role in Canavan disease. Cell Biosci. 2024;14(1):45. doi:10.1186/s13578-024-01224-6
41. Aspa Therapeutics. A combination retrospective medical history and prospective observational study of patients with Canavan disease (CANinform). ClinicalTrials.gov identifier NCT04126005. Completed 31 Aug 2025; updated 17 Apr 2026. Available from: https://clinicaltrials.gov/study/NCT04126005
42. Massachusetts General Hospital. A natural history study of Canavan disease: MGH site. ClinicalTrials.gov identifier NCT02851563. Completed Apr 2019. Available from: https://clinicaltrials.gov/study/NCT02851563
43. Karimzadeh P, Jafari N, Nejad Biglari H, Rahimian E, Ahmadabadi F, Nemati H, et al. The clinical features and diagnosis of Canavan's disease: a case series of Iranian patients. Iran J Child Neurol. 2014;8:66-71. PMID 25657773. Available from: https://pubmed.ncbi.nlm.nih.gov/25657773/
44. Masri A, Wahsh SA. Manifestations and treatment of epilepsy in children with neurometabolic disorders: a series from Jordan. Seizure. 2014;23:10-5. doi:10.1016/j.seizure.2013.08.006
45. Saudubray JM, García-Cazorla Á. Clinical Approach to Inborn Errors of Metabolism in Paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, sections 1.5.1.2 and 1.5.6, Tables 1.22, 1.25 and 1.28. doi:10.1007/978-3-662-63123-2_1
46. Yalcinkaya C, Benbir G, Salomons GS, Karaarslan E, Rolland MO, Jakobs C, et al. Atypical MRI findings in Canavan disease: a patient with a mild course. Neuropediatrics. 2005;36:336-9. doi:10.1055/s-2005-872878
47. Nguyen HV, Ishak GE. Canavan disease: unusual imaging features in a child with mild clinical presentation. Pediatr Radiol. 2015;45:457-60. doi:10.1007/s00247-014-3116-8
48. Sarret C, Boespflug-Tanguy O, Rodriguez D. Atypical clinical and radiological course of a patient with Canavan disease. Metab Brain Dis. 2016;31:475-9. doi:10.1007/s11011-015-9767-9
49. Kimiskidis VK, Papaliagkas V, Papagiannopoulos S, Zafeiriou D, Kazis D, Tsatsali-Foroglou E, et al. Investigation of the motor system in two siblings with Canavan's disease: a combined transcranial magnetic stimulation (TMS)-diffusion tensor imaging (DTI) study. Metab Brain Dis. 2017;32:307-10. doi:10.1007/s11011-017-9955-x
50. Tacke U, Olbrich H, Sass JO, Fekete A, Horvath J, Ziyeh S, et al. Possible genotype-phenotype correlations in children with mild clinical course of Canavan disease. Neuropediatrics. 2005;36:252-5. doi:10.1055/s-2005-865865
51. Delaney KE, Kralik SF, Hainline BE, Golomb MR. An atypical case of Canavan disease with stroke-like presentation. Pediatr Neurol. 2015;52:218-21. doi:10.1016/j.pediatrneurol.2014.10.019
52. Benson MD, Plemel DJA, Freund PR, Lewis JR, Sass JO, Bähr L, et al. Severe retinal degeneration in a patient with Canavan disease. Ophthalmic Genet. 2021;42:75-8. doi:10.1080/13816810.2020.1827441
53. Zafeiriou DI, Kleijer WJ, Maroupoulos G, Anastasiou AL, Augoustidou-Savvopoulou P, Papadopoulou F, et al. Protracted course of N-acetylaspartic aciduria in two non-Jewish siblings: identical clinical and magnetic resonance imaging findings. Brain Dev. 1999;21:205-8. PMID 10372908. Available from: https://pubmed.ncbi.nlm.nih.gov/10372908/
54. Firth HV, Hurst JA. Leukodystrophy/leukoencephalopathy. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 208–9.
55. Firth HV, Hurst JA. Developmental regression. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 128–9.
56. Firth HV, Hurst JA. Macrocephaly. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 224–5.
57. U.S. Food and Drug Administration. LIORESAL INTRATHECAL (baclofen injection), prescribing information. Version with effective date 9 Dec 2022, obtained through the openFDA API on 26 Oct 2026. Available from: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22LIORESAL%22
58. U.S. Food and Drug Administration. BOTOX (onabotulinumtoxinA) for injection, prescribing information. Allergan, Inc. Version with effective date 18 Nov 2023, obtained through the openFDA API on 26 Oct 2026. Available from: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22BOTOX%22
59. Health Resources and Services Administration. Recommended Uniform Screening Panel (core and secondary conditions as of January 2023) [Internet]. Rockville (MD): HRSA; page revised Jul 2024, accessed 26 Oct 2026. Available from: https://www.hrsa.gov/advisory-committees/heritable-disorders/rusp
60. Posern C, Dreyer B, Maier SL, Eichler F, Gelb MH, Santer R, et al. Quantification of N-acetyl-l-aspartate in dried blood spots: a simple and fast LC-MS/MS neonatal screening method for the diagnosis of Canavan disease. Mol Genet Metab. 2024;142(2):108489. doi:10.1016/j.ymgme.2024.108489
61. RTI International. Early Check: expanded screening in newborns. ClinicalTrials.gov identifier NCT03655223. Updated 1 Jul 2026. Available from: https://clinicaltrials.gov/study/NCT03655223
62. Surendran S, Bamforth FJ, Chan A, Tyring SK, Goodman SI, Matalon R. Mild elevation of N-acetylaspartic acid and macrocephaly: diagnostic problem. J Child Neurol. 2003;18:809-12. doi:10.1177/08830738030180111601
63. Touati G, Mochel F, Artuch R. Diagnostic Procedures. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Capítulo 3. doi:10.1007/978-3-662-63123-2_3
64. Matalon R, Kaul R, Michals K. Canavan disease: biochemical and molecular studies. J Inherit Metab Dis. 1993;16:744-52. doi:10.1007/BF00711906
65. Matalon RM, Michals-Matalon K. Spongy degeneration of the brain, Canavan disease: biochemical and molecular findings. Front Biosci. 2000;5:D307-11. doi:10.2741/matalon
66. Matalon R, Michals K, Kaul R. Canavan disease: from spongy degeneration to molecular analysis. J Pediatr. 1995;127:511-7. doi:10.1016/s0022-3476(95)70105-2
67. Rahimian E, Tahsini MR, Karimzadeh P, Khalilian S, Fathi M, Miryounesi M, et al. Aspartoacylase (ASPA) gene mutations and neuroimaging features in Iranian patients with Canavan disease: a descriptive study. Brain Dev. 2026;48(4):104564. doi:10.1016/j.braindev.2026.104564
68. Çakar NE, Aksu Uzunhan T. A case of juvenile Canavan disease with distinct pons involvement. Brain Dev. 2020;42:222-5. doi:10.1016/j.braindev.2019.11.009
69. Jauhari P, Saini L, Chakrabarty B, Kumar A, Gulati S. Juvenile Canavan disease: a leukodystrophy without white matter changes. Neuropediatrics. 2018;49:420-1. doi:10.1055/s-0038-1672175
70. Tahsini MR, Wylezinska-Arridge M, Zare M, Saket S, Farahvash MA, Morovvati M, et al. Diagnostic value of quantitative MR spectroscopy metabolite ratios for Canavan disease. Front Radiol. 2026;6:1826170. doi:10.3389/fradi.2026.1826170
71. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. L-2-Hydroxyglutaric aciduria. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 95–7.
72. SyriMed. Baclofen 10 mg/5 ml oral solution: summary of product characteristics [Internet]. Electronic medicines compendium (emc), United Kingdom; text revised 10 Jul 2026, accessed 26 Oct 2026. Available from: https://www.medicines.org.uk/emc/product/11850/smpc
73. U.S. Food and Drug Administration. OZOBAX DS (baclofen) oral solution, prescribing information. Version with effective date 13 Mar 2026, obtained through the openFDA API on 26 Oct 2026. Available from: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22OZOBAX%22
74. U.S. Food and Drug Administration. Levetiracetam tablets and oral solution, prescribing information. Version with effective date 21 Nov 2025, obtained through the openFDA API on 26 Oct 2026. Available from: https://api.fda.gov/drug/label.json?search=openfda.generic_name:%22LEVETIRACETAM%22
75. European Medicines Agency. Keppra (levetiracetam): EPAR, product information, summary of product characteristics, sections 4.1–4.2 [Internet]. Amsterdam: EMA; accessed 26 Oct 2026. Available from: https://www.ema.europa.eu/en/documents/product-information/keppra-epar-product-information_en.pdf
76. Myrtelle Inc. Phase 1/2, open label, sequential cohort study of a single intracranial dose of AVASPA gene therapy for treatment of children with typical Canavan disease. ClinicalTrials.gov identifier NCT04833907. Updated 14 Nov 2025. Available from: https://clinicaltrials.gov/study/NCT04833907
77. Madhavarao CN, Arun P, Anikster Y, Mog SR, Staretz-Chacham O, Moffett JR, et al. Glyceryl triacetate for Canavan disease: a low-dose trial in infants and evaluation of a higher dose for toxicity in the tremor rat model. J Inherit Metab Dis. 2009;32(5):640-50. doi:10.1007/s10545-009-1155-3
78. Segel R, Anikster Y, Zevin S, Steinberg A, Gahl WA, Fisher D, et al. A safety trial of high dose glyceryl triacetate for Canavan disease. Mol Genet Metab. 2011;103(3):203-6. doi:10.1016/j.ymgme.2011.03.012
79. Sheba Medical Center. Phase 1 treatment with GTA in two infants with Canavan disease. ClinicalTrials.gov identifier NCT00278707 (and related registration NCT00724802, Shaare Zedek Medical Center). Unknown status. Available from: https://clinicaltrials.gov/study/NCT00278707
80. Francis JS, Markov V, Leone P. Dietary triheptanoin rescues oligodendrocyte loss, dysmyelination and motor function in the nur7 mouse model of Canavan disease. J Inherit Metab Dis. 2014;37(3):369-81. doi:10.1007/s10545-013-9663-6
81. Janson CG, Assadi M, Francis J, Bilaniuk L, Shera D, Leone P. Lithium citrate for Canavan disease. Pediatr Neurol. 2005;33(4):235-43. doi:10.1016/j.pediatrneurol.2005.04.015
82. Assadi M, Janson C, Wang DJ, Goldfarb O, Suri N, Bilaniuk L, Leone P. Lithium citrate reduces excessive intra-cerebral N-acetyl aspartate in Canavan disease. Eur J Paediatr Neurol. 2010;14(4):354-9. doi:10.1016/j.ejpn.2009.11.006
83. Assistance Publique – Hôpitaux de Paris. Evaluation of the tolerance and efficiency of a combined oral therapy with lithium and GTA in patients with Canavan disease. ClinicalTrials.gov identifier NCT00657748. Withdrawn; updated 21 Apr 2015. Available from: https://clinicaltrials.gov/study/NCT00657748
84. Nešuta O, Thomas AG, Alt J, Hin N, Neužilová A, Long S, et al. High throughput screening cascade to identify human aspartate N-acetyltransferase (ANAT) inhibitors for Canavan disease. ACS Chem Neurosci. 2021;12(18):3445-55. doi:10.1021/acschemneuro.1c00455
85. Poddar NK, Wijayasinghe YS, Viola RE. Identification of potential pharmacological chaperones that selectively stabilize mutated aspartoacylases in Canavan disease. Biochim Biophys Acta Proteins Proteom. 2024;1872(6):141043. doi:10.1016/j.bbapap.2024.141043
86. McPhee SW, Janson CG, Li C, Samulski RJ, Camp AS, Francis J, et al. Immune responses to AAV in a phase I study for Canavan disease. J Gene Med. 2006;8(5):577-88. doi:10.1002/jgm.885
87. Leone P, Shera D, McPhee SW, Francis JS, Kolodny EH, Bilaniuk LT, et al. Long-term follow-up after gene therapy for Canavan disease. Sci Transl Med. 2012;4(165):165ra163. doi:10.1126/scitranslmed.3003454
88. Janson CG, McPhee SW, Francis J, Shera D, Assadi M, Freese A, et al. Natural history of Canavan disease revealed by proton magnetic resonance spectroscopy (1H-MRS) and diffusion-weighted MRI. Neuropediatrics. 2006;37(4):209-21. doi:10.1055/s-2006-924734
89. Aspa Therapeutics. A study of AAV9 gene therapy in participants with Canavan disease (CANaspire clinical trial). ClinicalTrials.gov identifier NCT04998396. Updated 17 Apr 2026. Available from: https://clinicaltrials.gov/study/NCT04998396
90. BridgeBio Pharma, Inc. Quarterly report on Form 10-Q for the quarter ended June 30, 2026. Washington (DC): U.S. Securities and Exchange Commission; filed 10 Aug 2026. Accession no. 0001743881-26-000027. Available from: https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001743881&type=10-Q
91. University of Florida. Expanded access trial of systemic delivery of aspartoacylase ASPA (rAAV9-CB6-AspA) gene vector in a single patient with Canavan disease. ClinicalTrials.gov identifier NCT05317780. No longer available; updated 23 Aug 2023. Available from: https://clinicaltrials.gov/study/NCT05317780
92. Aldagen. ALD-101 adjuvant therapy of unrelated umbilical cord blood transfusion (UCBT) in patients with inherited metabolic diseases. ClinicalTrials.gov identifier NCT00654433. Terminated; updated 8 Jul 2014. Available from: https://clinicaltrials.gov/study/NCT00654433
93. Feng L, Chao J, Tian E, Li L, Ye P, Zhang M, et al. Cell-based therapy for Canavan disease using human iPSC-derived NPCs and OPCs. Adv Sci (Weinh). 2020;7(23):2002155. doi:10.1002/advs.202002155
94. Chao J, Feng L, Ye P, Chen X, Cui Q, Sun G, et al. Therapeutic development for Canavan disease using patient iPSCs introduced with the wild-type ASPA gene. iScience. 2022;25(6):104391. doi:10.1016/j.isci.2022.104391
95. Jackson N, Feng L, Chao J, Ye P, Luong Q, Galo K, et al. Human iPSC-derived neural progenitor cells rescue motor function and brain pathology in symptomatic Canavan disease mice. Stem Cell Reports. 2026;21(2):102785. doi:10.1016/j.stemcr.2025.102785
96. European Commission. Community Register of orphan medicinal products for human use: active designations EU/3/20/2287 and EU/3/22/2729 [Internet]. Brussels: European Commission; accessed 26 Oct 2026. Available from: https://ec.europa.eu/health/documents/community-register/html/reg_od_act.htm
97. Townsend EL, Kiefer M, Leiro B, Gillan E, Marion J, Khan A, Tavakkoli F. Measurement of motor function in children with Canavan disease: concordance between remote and in-person assessments. Pediatr Neurol. 2026;178:110-7. doi:10.1016/j.pediatrneurol.2026.02.008
98. Boitsios G, Aeby A, Andrei AM. Case 336: Canavan disease. Radiology. 2025;315(1):e240184. doi:10.1148/radiol.240184
99. Gregg AR, Aarabi M, Klugman S, Leach NT, Bashford MT, Goldwaser T, et al. Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(10):1793-806. doi:10.1038/s41436-021-01203-z
100. American College of Obstetricians and Gynecologists, Committee on Genetics. Committee Opinion No. 691: Carrier screening for genetic conditions. Obstet Gynecol. 2017 (março de 2017; reafirmado em 2025). doi:10.1097/AOG.0000000000001952
101. Kronn D, Oddoux C, Phillips J, Ostrer H. Prevalence of Canavan disease heterozygotes in the New York metropolitan Ashkenazi Jewish population. Am J Hum Genet. 1995;57:1250-2. PMID 7485179. Available from: https://pubmed.ncbi.nlm.nih.gov/7485179/
102. Fares F, Badarneh K, Abosaleh M, Harari-Shaham A, Diukman R, David M. Carrier frequency of autosomal-recessive disorders in the Ashkenazi Jewish population: should the rationale for mutation choice for screening be reevaluated? Prenat Diagn. 2008;28:236-41. doi:10.1002/pd.1943
103. Matalon R, Michals-Matalon K. Molecular basis of Canavan disease. Eur J Paediatr Neurol. 1998;2:69-76. doi:10.1016/s1090-3798(98)80044-5
104. Elpeleg ON, Shaag A. The spectrum of mutations of the aspartoacylase gene in Canavan disease in non-Jewish patients. J Inherit Metab Dis. 1999;22:531-4. doi:10.1023/a:1005512524957
105. Kiefer M, Khan A, Leiro B, Yavorsky C, Laforet G, Kirby K, Townsend E. Feasibility, acceptability, and reliability of remote motor assessment in children with Canavan disease. Pediatr Neurol. 2025;164:129-36. doi:10.1016/j.pediatrneurol.2025.01.006