Leukodystrophies — General Approach (Including LBSL)

Genetic diseases primarily affecting the white matter of the central nervous system: definition, classification, magnetic resonance imaging recognition, genomic strategy, treatable causes, supportive care and approved therapies; leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation (LBSL, DARS2) addressed in dedicated subsections.

Literature and regulatory information updated: 9 October 2026.

0. Initial Reference Table

1. Introduction

1.1. From “diffuse sclerosis” to the genomic concept

The history of leukodystrophies began more than a century ago, when Pelizaeus and Merzbacher described a family with “diffuse sclerosis” and absence of myelin, the disorder now called Pelizaeus-Merzbacher disease (PMD) [2]. The term “leukodystrophy” was first used in 1928, in relation to metachromatic leukodystrophy (MLD), to describe inherited progressive diseases with white matter (WM) degeneration [2]. Until the 1980s, a concept centred on myelin and oligodendrocytes prevailed: leukodystrophies were progressive genetic diseases primarily affecting myelin sheath synthesis or maintenance [2]. Magnetic resonance imaging (MRI) subsequently transformed the field by enabling recognition of lesion distribution patterns that proved characteristic of distinct diseases; the organisation of these patterns into an algorithm, proposed by Schiffmann and van der Knaap in 2009, became the basis of the radiological reasoning that still guides diagnosis [2, 29]. Over the last two decades, genes have been identified first through linkage analysis and then through exome and genome sequencing, and many have proved to be involved in general cellular (housekeeping) processes, such as cytoplasmic and mitochondrial translation, RNA polymerase III transcription or control of the cellular stress response, making a myelin-centred definition too narrow [2, 30].

The scale of the diagnostic problem changed in parallel. The 2014 GeneReviews overview chapter, retired in January 2020 because it covered too broad a phenotype, and the 2015 consensus documents reported that around half of children with presumed genetic WM disease remained without a specific diagnosis, even after high-quality MRI and a complete biochemical battery [1, 9, 14]. The second edition of the Oxford Desk Reference (2017) contains two apparently contradictory statements, “up to 80%” with a molecular diagnosis and “around 50%” without an aetiology, reflecting different periods: the latter describes the pre-exome era [31]. Prospective exome and genome cohorts subsequently documented molecular diagnostic rates close to 80% and a reduction in time to diagnosis from more than five years to less than six months after MRI in many families [6, 32].

1.2. GLIA case definition (2015)

The Global Leukodystrophy Initiative (GLIA), an international consortium founded in 2013, published a case definition in 2015, established using a modified Delphi method with 13 experts who reviewed 91 inherited diseases with WM abnormalities [1, 33]. According to this definition, leukodystrophies are inherited diseases affecting CNS WM, with or without peripheral nervous system involvement, which share abnormalities of glial cells or the myelin sheath; where neuropathology is known, oligodendrocytes, astrocytes and other non-neuronal cells are primarily involved, although significant axonal pathology is suspected in some [1]. On MRI, affected WM is T2-hyperintense; T1 signal that is mildly hypointense, isointense or hyperintense relative to the cortex is compatible with hypomyelinating leukodystrophy, whereas clearly hypointense T1 signal is compatible with demyelinating leukodystrophy [1]. Acquired myelin diseases (multiple sclerosis and related disorders), non-genetic infectious, postinfectious, toxic and vascular lesions, diseases primarily affecting cortical neurons and grey matter, and inborn errors of metabolism dominated by systemic manifestations (liver, muscle, heart), even if they affect WM, are excluded [1].

The consensus distinguished leukodystrophies, of which it listed more than 30, from genetic leukoencephalopathies (gLE), of which it listed more than 61; all leukodystrophies are gLE, but not all gLE are leukodystrophies [1]. The criterion placing a disease among gLE is evidence of primary neuronal involvement or systemic manifestations overshadowing those of WM: intramyelinic oedema in maple syrup urine disease (MSUD) is systemic in origin and is a gLE, whereas intramyelinic oedema in ClC-2 and GlialCAM disorders and in megalencephalic leukoencephalopathy with subcortical cysts (MLC), resulting from an astrocytic defect, defines leukodystrophies [1]. The definition does not discriminate by organelle: lysosomal (Krabbe, MLD, Salla disease, fucosidosis), peroxisomal (X-linked ALD, X-ALD, and peroxisomal biogenesis disorders), mitochondrial (LBSL, leukoencephalopathy with thalamus and brain stem involvement and high lactate, LTBL) and intermediary metabolic (Canavan, polyglucosan body disease, cerebrotendinous xanthomatosis, CTX) leukodystrophies coexist with gLE affecting the same organelles [1]. Nor does it require progression, since it included remitting MLC caused by a dominant HEPACAM variant, the very slowly progressive form of 4H leukodystrophy and adult forms of Alexander disease [1]. Hereditary vasculopathies (COL4A1/COL4A2, CADASIL, cerebroretinal microangiopathy with calcifications and cysts, CRMCC) were excluded because of their multifocal pattern, whereas Aicardi-Goutières syndrome (AGS) was included with 91.7% agreement in the first round, because of its diffuse pattern with calcifications and the role of astrocytic interferon [1]. 3-Phosphoglycerate dehydrogenase deficiency and MCT8 transporter deficiency (SLC16A2) did not achieve full consensus and were classified as gLE by the majority [1].

The authors themselves emphasised the limitations: neuropathological data are lacking for several diseases, TUBB4A-associated disease may prove to be primarily neuronal, and individual patients within recognised spectra, such as PLP1-related spastic paraparesis, brain stem-predominant Alexander disease or adrenomyeloneuropathy (AMN), do not fulfil the radiological criteria; the definition is intended for epidemiological classification of diseases rather than clinical classification of individuals [1]. Table 1.1 summarises the consensus list of leukodystrophies and the percentage agreement in the first round.

Table 1.1. Leukodystrophies according to the 2015 GLIA consensus (percentage agreement in the first round; 100% indicates immediate consensus) [1]

1.3. The broad definition and cellular classification (2017)

In 2017, van der Knaap and Bugiani proposed a broader definition whereby leukodystrophies encompass all genetically determined disorders primarily affecting CNS WM, irrespective of the structural component involved, molecular process or course, which may be progressive, static, episodic or improving [2]. This definition differs from GLIA, which excludes primary neuronal diseases, vasculopathies and systemic inborn errors, and the two coexist in the literature: the GLIA definition serves epidemiological delineation, and the 2017 definition serves mechanistic research [1, 2]. The classical four-category classification (hypomyelinating, demyelinating, dysmyelinating and myelinolytic or vacuolating) has been replaced by a cellular pathology classification, presented in section 2 (Table 2.1), which groups myelin disorders, astrocytopathies, leuko-axonopathies, microgliopathies and leuko-vasculopathies [2]. LBSL is classified among the leuko-axonopathies [2].

1.4. Hypomyelination and demyelination

The radiological distinction between hypomyelination and other WM pathologies remains the first decision point in the diagnostic algorithm [9, 29]. Hypomyelination denotes deficient myelin production and is defined by an unchanged pattern of deficient myelination on two MRI scans at least six months apart in a child older than one year; T2 signal is increased but less bright than in demyelinating disorders, and WM may appear relatively normal on T1 [9, 31]. Demyelinating or destructive disorders, by contrast, involve loss of or damage to previously formed myelin and show marked T1 hypointensity [1, 14]. Hypomyelinating disorders currently constitute the largest group of undiagnosed leukodystrophies because their MRI pattern is subtler and radiological diagnosis more difficult [31].

1.5. LBSL: founding description and nomenclature

In 2003, van der Knaap and colleagues described eight patients with a previously unrecognised MRI pattern, featuring heterogeneous cerebral WM abnormalities and selective involvement of brain stem and spinal cord tracts, with increased lactate in abnormal WM on spectroscopy; the clinical course was slowly progressive, with pyramidal, cerebellar and posterior column signs, and inheritance was probably autosomal recessive [5]. The disease name, leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation, describes these three elements [5]. In 2007, Scheper and colleagues used linkage mapping and shared haplotypes to identify variants in DARS2, encoding mitochondrial aspartyl-tRNA synthetase, in all 30 families studied [4]. The lactate elevation in the name refers to the spectroscopy signal in abnormal WM rather than systemic lactic acidosis; blood and cerebrospinal fluid (CSF) lactate is normal in most patients or only mildly and inconsistently elevated [3, 8, 21]. The description of a severe early infantile form and, in 2025, axonal Charcot-Marie-Tooth (CMT) neuropathy without leukoencephalopathy broadened the spectrum into a continuum between the CNS and peripheral nervous system [34, 35].

1.6. Scope of this article

This article covers the general approach to leukodystrophies, with LBSL as the entity developed in dedicated subsections. X-ALD, MLD, Krabbe disease and AGS appear here only in the diagnostic algorithm, the differential diagnosis and the table of approved therapies. The other leukodystrophies (Alexander, Canavan, VWM, MLC, PMD, POLR3, ADLD, H-ABC, HCC and others) appear as examples of patterns and in the differential diagnosis, with emphasis on Alexander disease, which in September 2026 obtained the first approved therapy [22].

2. Molecular Genetics and Pathophysiology

2.1. Genetic architecture of the group

Leukodystrophies are inherited through all Mendelian and mitochondrial mechanisms: autosomal recessive, the most frequent; autosomal dominant, often through a de novo variant (Alexander disease due to GFAP, TUBB4A-related disease); X-linked (X-ALD due to ABCD1, PMD due to PLP1); and mitochondrial [9, 14]. The minimum gene list proposed by GLIA in 2015 already included, alongside genes for classical leukodystrophies, dozens of genes for gLE, mitochondrial diseases and hereditary spastic paraplegias, and heterogeneity has continued to grow with exome and genome sequencing [9, 30]. In North American cohorts of patients remaining undiagnosed after standard assessment, only around 20% of diagnoses obtained by next-generation sequencing (NGS) involved “canonical” leukodystrophy genes, with the majority corresponding to gLE, mitochondrial diseases or neurodevelopmental disorders with WM abnormalities [32, 36]. This finding has direct implications for test selection, discussed in section 5.

Molecular mechanisms span almost the whole of cell biology. Defects affect structural myelin proteins (PLP1), lysosomal enzymes (ARSA, GALC), peroxisomal transporters (ABCD1), astrocytic proteins (GFAP, MLC1, HEPACAM, CLCN2), translation factors (EIF2B1–EIF2B5), cytoplasmic and mitochondrial aminoacyl-tRNA synthetases (DARS1, DARS2, EARS2, AARS2, RARS1), RNA polymerase III subunits (POLR3A, POLR3B, POLR1C), the cytoskeleton (TUBB4A), the nuclear lamina (LMNB1), the microglial receptor CSF1R and intermediary metabolic enzymes (ASPA, CYP27A1) [1, 2, 9]. Table 3 of the GLIA consensus illustrates that subcellular localisation of the defect does not determine classification: what matters is whether the defect is intrinsic to glial cells (leukodystrophy) or the toxin is generated systemically (gLE), which is why X-ALD, Canavan disease and peroxisomal biogenesis disorders are leukodystrophies, whereas phenylketonuria and homocystinuria are not, since their WM abnormalities regress with appropriate treatment [1]. The source itself contains an internal inconsistency, classifying L-2-hydroxyglutaric aciduria as a leukodystrophy in Table 3 and as a gLE in the text and Table 2 [1].

2.2. Classification by cellular pathology

The van der Knaap and Bugiani classification groups diseases by the cell type or structure primarily affected; a disease may belong to more than one category, and assignment also relies on imaging and protein function when neuropathology is unavailable [2]. The original text announces six categories but lists five, with three subtypes of myelin disorders [2].

Table 2.1. Classification of leukodystrophies by cellular pathology (examples, non-exhaustive) [2]

The authors draw a therapeutic implication from this classification: knowing the cell types involved is essential for therapy design, and no leukodystrophy is likely to be cured by a single treatment; combined approaches will be required (cell replacement, drugs and small molecules, endogenous repair) [2]. Mechanisms of WM injury include hypomyelination, dysmyelination, demyelination, intramyelinic oedema (microvacuolation), interstitial oedema, microgliosis and astrogliosis, axonal loss, and rarefaction or cystic degeneration [6].

2.3. Illustrative gene-by-gene mechanisms

Table 2.2 summarises, for the leukodystrophies used as examples in this article, the gene, the mechanism and the genetic features of diagnostic relevance.

Table 2.2. Genetics and mechanisms of selected leukodystrophies

In POLR3 leukodystrophy, POLR3A variants are associated with later onset and faster progression, and POLR3B variants with earlier onset, a slower course, more cerebellar atrophy and preserved myelination of corticospinal tracts in the internal capsules; the common POLR3B NM_018082.6:c.1568T>A p.(Val523Glu) variant is very mild, and homozygotes are asymptomatic or very mildly symptomatic in early adulthood [10]. Biallelic hypomorphic POLR3A variants cause progressive adolescent-onset spastic ataxia, and biallelic truncating variants cause neonatal progeroid Wiedemann-Rautenstrauch syndrome, already treated as an established entity in 2022, although the 2017 GeneReviews presented it as requiring replication [10, 37]. In a study of 618 index cases with hereditary spastic paraplegia or cerebellar ataxia, around 3.1% had compound heterozygous POLR3A variants, a diagnostic yield rather than a prevalence figure [8]. Biallelic POLR1C variants distinct from leukodystrophy variants cause Treacher Collins syndrome [10]. In ADLD, upstream deletions are associated with earlier onset, without initial dysautonomia, and less brain stem and cerebellar involvement than duplication; de novo missense variants in LMNB1 cause a different allelic disorder with primary microcephaly, relative short stature and intellectual disability [11]. In Canavan disease, the ASPA NM_000049.4:c.854A>C p.(Glu285Ala) variant accounts for 80–85% of alleles in Ashkenazi Jews and c.693C>A p.(Tyr231Ter) for around 13%; c.914C>A p.(Ala305Glu) predominates in non-Ashkenazi Europeans, and an intronic insertion of around 2600 bp, detectable by long-read sequencing, may be the most common pathogenic variant across several ancestries [31, 40, 41].

2.4. DARS2 and LBSL pathophysiology

The DARS2 gene (1q25.1; NCBI Gene 55157; 18 exons; MANE Select NM_018122.5, protein NP_060592.2) encodes mitochondrial aspartate-tRNA ligase (mtAspRS), a class II aminoacyl-tRNA synthetase that charges tRNA-Asp in mitochondria and, by homology with the bacterial enzyme, is thought to function as a homodimer [3, 7]. GeneReviews prints the transcript as “NM_108122.5” and locates the founder haplotype on “chromosome 1p25”; both are typographical errors, since the gene RefSeq is NM_018122 and the cytogenetic location is 1q25.1 [7, 8].

The mechanism is partial loss of function. Almost all patients have, on one allele, a variant affecting exon 3 splicing, located upstream of this exon in a stretch of 10 to 11 cytosines around 10 nucleotides from the exon boundary; these variants cause exon 3 skipping, with a frameshift and mRNA degradation by nonsense-mediated decay, but reduce rather than abolish correct splicing, leaving a small amount of normal protein (leaky variants) [3]. Complete absence of activity is presumed incompatible with life, explaining the near absence of homozygotes for null alleles and the occurrence of severe prenatal and infantile forms in patients without the common intronic variant, carrying more severe loss-of-function variants [3, 46]. Several missense variants markedly reduce aminoacylation by the recombinant protein, and mtAspRS activity is markedly reduced in patients' cells [4, 19].

A paradoxical feature is the normal respiratory chain. Scheper and colleagues found normal mitochondrial complex activities in fibroblasts and lymphoblasts, and GeneReviews states that muscle biopsies, fibroblasts and lymphoblasts show no mitochondrial dysfunction [3, 4]. Saudubray's textbook classifies DARS2 among tRNA charging genes with a typical combined defect of complexes I, III, IV and V, corresponding to the gene's functional class rather than the usual findings in accessible tissues of patients with LBSL [3, 47]. Selective vulnerability of long brain stem and spinal cord tracts, increased fractional anisotropy with reduced apparent diffusion coefficient in WM, and vacuolar changes with myelin separation described in two siblings with severe disease suggest a process starting in the axon; the 2026 preclinical work describes LBSL as a primary axonopathy with secondary demyelination [2, 3, 27].

Table 2.3. Clinically relevant DARS2 variants, normalised to NM_018122.5 (GRCh38)

Family LD_0315 in Helman's 2020 study has an intronic variant described as “c.228-21delTinsCC”, which in VariantValidator corresponds to a different variant from the common one (NM_018122.5:c.228-21delinsCC); whether this is an alternative spelling of the common variant could not be determined, so it was not normalised [32, 48]. In the 2026 Norwegian case, variants were described against NM_018122.4; normalisation to the MANE version (.5) does not change the c. nomenclature [46, 48].

2.5. Genotype–phenotype correlations, allelic diseases and penetrance in LBSL

In the cross-sectional study of the Amsterdam and Helsinki databases (78 patients with two variants, clinical data for 66), 60 different variants were found, including 13 novel variants; 94% of patients had an intron 2 splice variant, and all were compound heterozygotes except 4 of 42 cases published by other groups [19]. Some combinations were consistently associated with a mild phenotype, but groups were too small for formal correlation; prenatal and early infantile forms lack the common leaky splice variants and are associated with more severe loss-of-function variants [3, 19]. Intrafamilial variability can be considerable: in a Sudanese family, siblings with the same variants ranged from an apparently healthy individual with brisk reflexes as the only finding at age 20 to disabled patients [3, 46].

Allelic diseases include autosomal recessive axonal CMT (OMIM 621485), described in 2025 in five individuals from three families in Spain, France and Estonia, with childhood-onset progressive sensorimotor neuropathy, distal weakness and foot deformities, without leukoencephalopathy; the genotypes (NM_018122.5) combined the hypomorphic c.713C>T p.(Ser238Phe) with c.1006C>T p.(Arg336Cys) or with the variant published as p.(Ile25ThrfsTer38) (published designation; c. not mapped), and c.492+2T>C with c.1508C>T p.(Pro503Leu), which reduces enzyme activity by 75%; functional studies confirmed reduced activity and dimerisation, and the authors proposed a continuum between CNS and peripheral disease and DARS2 screening in CMT cohorts [3, 35, 50]. Adult-onset presentations mimicking isolated hereditary spastic paraplegia, with proprioceptive dysfunction and no cognitive changes, have also been described, as has, with a homozygous variant, exercise-induced episodic ataxia responsive to acetazolamide [3]. The c. nomenclature of the missense variants in the CMT study is not given in the abstract consulted and was not normalised.

Penetrance in leukodystrophies generally is under reassessment. A Utah genomic estimate summed allele frequencies of 4564 pathogenic variants in 49 genes and predicted an overall incidence of 1/4733 live births, with values for GALC 8.3 times higher than published estimates; the authors acknowledge that the estimate includes variants of uncertain penetrance and late-onset forms [18]. In a population study of 210,983 mostly healthy individuals in the Intermountain health system (US), released in 2026 as a non-peer-reviewed preprint, nine individuals had pathogenic genotypes in CSF1R, PLP1, POLR3A and SNORD118 without a clinical diagnosis or characteristic MRI, suggesting variable penetrance and expressivity with implications for genomic newborn screening and variant interpretation [51]. ADLD penetrance is considered complete, with onset in the fourth or fifth decade and equal in both sexes [11].

3. Clinical Spectrum and Manifestations by System

3.1. Core neurological manifestations

Leukodystrophies almost always affect the motor system, distinguishing them from primary neuronal diseases dominated by cognitive decline and seizures [9]. In infants, hypotonia evolves into spasticity ranging from mild diplegia to severe tetraplegia, accompanied by impaired swallowing, chewing and sometimes breathing, and orthopaedic complications such as scoliosis and dislocation of major joints [14]. Spasticity may mask dystonia and dyskinesias, which are prominent, for example, in MCT8 deficiency and TUBB4A-related disease; ataxia predominates and is disabling in VWM and 4H leukodystrophy [10, 14]. Cognitive decline is generally much less pronounced than motor decline, and dementia is not early, with exceptions such as juvenile and adult X-ALD and MLD, where declining school performance or behavioural changes may herald disease [9, 14].

The temporal pattern is informative. In infants, motor delay is more typical of hypomyelinating disorders and motor regression of demyelinating or destructive disorders; in children, falls and clumsy gait appear; in adolescents and adults, loss of functional abilities, such as sporting abilities [9]. Some leukodystrophies are so slowly progressive that they are mistaken for static encephalopathy, and a minority improve slowly, such as MLC due to dominant HEPACAM and LTBL [9]. Acute deterioration after intercurrent illness or minor head trauma is documented in VWM and has been reported in LBSL, PMD, X-ALD and ADLD [3, 11, 14, 38]. Signs include pyramidal signs (spastic tetraparesis or paraparesis), extrapyramidal signs (dystonia, chorea), cerebellar and sensory signs (ataxia), dysautonomia, altered head circumference, seizures, extreme early irritability, school decline, new-onset hyperactivity and personality change [9].

3.2. Seizures and epilepsy

Seizures are usually late, except in Alexander disease, where they frequently present first [14]. The 2017 GLIA care consensus reported seizures in almost 50% of patients, a figure derived from the Utah cohort (epilepsy in 49%) [16, 33]. A retrospective cohort of 704 children and adolescents with leukodystrophy at 8 US paediatric centres (2013–2020), including children identified through newborn screening, found at least one seizure in 29%, and epilepsy in 80% of those who had a seizure; electroencephalography (EEG) was abnormal in 84% of patients with seizures and 54% of the remainder [52]. In multivariable analysis, each year of age increased the risk (odds ratio, OR, 1.13; 95% CI 1.07–1.19), as did generalised epileptiform discharges (OR 13.4; 2.22–259), focal or multifocal discharges (OR 3.56; 2.04–6.36) and focal slowing (OR 3.57; 1.58–8.58), whereas diffuse slowing and abnormal MRI were not predictors [52]. Astrocytopathies had more seizures than myelinopathies (41% versus 28%; p=0.046), and several leukodystrophies required two or more antiseizure medications [52]. The difference between 29% and “almost 50%” reflects different populations and periods: a recent multicentre paediatric cohort receiving follow-up and treatment versus a regional hospital cohort from 1999–2007 [16, 52].

3.3. Extraneurological manifestations

Signs outside the CNS often enable the diagnosis to be narrowed down, and the GLIA diagnostic consensus organised them by system [9]. Table 3.1 brings these together with findings from GeneReviews and the Queen Square adult review.

Table 3.1. Informative extraneurological signs

3.4. Adult forms

Adult-onset leukodystrophies differ from paediatric forms through a differential diagnosis dominated by acquired causes, a lower genetic diagnostic yield and their own aetiological spectrum [6, 21]. In the Queen Square group (London), around half (54) of 116 investigated patients received a diagnosis; the most frequent were small vessel disease, CSF1R disease, mitochondrial diseases, X-ALD and multiple sclerosis, while the remainder were highly heterogeneous, with 15 single-patient diagnoses [21]. CSF1R disease accounts for around 10% of adult leukodystrophies, and AARS2 leukoencephalopathy has a similar phenotype, recessive inheritance, mean onset at 29 years (15–44) and total dependence within around five years [21]. Adult clinical clues include premature ovarian failure, parkinsonism (CSF1R), chorea (Gordon Holmes), palatal myoclonus (Alexander), axonal neuropathy (X-ALD, LBSL, adult polyglucosan body disease, CTX), demyelinating neuropathy (MLD), xanthomas, early urinary frequency (polyglucosan body disease), early dysautonomia (ADLD), prominent early ataxia (CTX, Gordon Holmes, CLCN2), migraine with aura (CADASIL) and occasionally rapid progression within less than one year (X-ALD, Krabbe) [21]. Psychiatric symptoms are frequent in adult forms of MLD, Krabbe, Alexander and CSF1R disease [33].

3.5. Natural history of the group

The Utah regional children's hospital cohort (1999–2007) is the main source of natural history data for the group as a whole: among 122 children with inherited leukodystrophy, 49% had epilepsy, 43% a gastrostomy, 32% developmental regression and 15% endocrine abnormalities; 34% died, with a mean age at death of 8.2 years, and 51% had no final diagnosis [16]. The most frequent diagnoses were MLD (8.2%), PMD (7.4%), mitochondrial diseases (4.9%) and X-ALD (4.1%), and mean annual medical cost per patient was 22,579 dollars [16]. These figures predate exome sequencing and gene therapies, include gLE and undiagnosed cases, and should not be read as the current prognosis of a specific disease [16, 33]. Natural history has changed for forms treated by HSCT or gene therapy: in the United States, the first children with early-onset MLD treated with gene therapy now survive into their mid-teens, whereas those not treated before symptoms die at around 5–6 years [54].

3.6. LBSL — clinical features

LBSL is characterised by slowly progressive cerebellar ataxia, spasticity and posterior column dysfunction, with reduced position and vibration sense, more in the lower than the upper limbs; patients have increasing difficulty walking in the dark, manual dexterity is variably affected and tendon reflexes are preserved [3]. Motor deterioration usually begins in childhood or adolescence and sometimes in infancy or adulthood; dysarthria develops over time, but swallowing is generally preserved in childhood and adult forms [3]. Other manifestations are epilepsy, learning difficulties, cognitive decline and, after minor head trauma, reduced consciousness, neurological deterioration and fever, with only partial recovery [3]. Table 3.2 presents manifestation frequencies according to GeneReviews, based on more than 150 individuals and large published series, without prospective natural history data [3].

Table 3.2. Frequency of manifestations in LBSL [3]

The spectrum has three forms. The prenatal or early infantile form presents with almost invariable and often severe microcephaly, severe delay with no or very few motor milestones, severe cognitive impairment, almost invariable and frequently severe epilepsy, and frequent early death; in the review of around 3000 MRI scans of unknown leukoencephalopathies that identified it, the earliest onset was neonatal and the earliest death occurred at 20 months [3, 34]. The childhood-onset form is the most common and progresses slowly; according to GeneReviews, most become partly or fully wheelchair-dependent during adolescence, in their twenties or later [3]. The abstract of the 66-patient study, in turn, states that full wheelchair dependence is rare and generally occurs in adulthood; the two formulations refer to different degrees of dependence [3, 19]. The adult-onset form progresses slowly with mild impairment, and no patients are known to have become wheelchair-dependent [3].

In the sensory domain, severe posterior column dysfunction produces sensory ataxia; some patients have axonal neuropathy, but neurophysiology is frequently normal [3]. The most common presentation in adult series is childhood onset with distal weakness and wasting and pes cavus, which may be labelled hereditary sensorimotor neuropathy, followed by pyramidal signs and cerebellar and posterior column ataxia [21]. Most patients have normal intellectual ability, with some early learning difficulties and possible slow decline [3]. Routine laboratory tests and CSF are usually normal, with mildly and inconsistently raised blood or CSF lactate in a few patients [3].

3.7. Other leukodystrophies used as reference disorders

Table 3.3 summarises the clinical presentation of other leukodystrophies, for comparison; the full differential diagnosis is in section 6.

Table 3.3. Clinical features of selected leukodystrophies

4. Clinical Recognition and Warning Signs

4.1. When to suspect

Leukodystrophy is usually suspected when MRI requested for a neurologically abnormal child shows WM abnormalities, but suspicion should precede imaging in several clinical settings [9, 31]. In infants, typical situations are motor delay with nystagmus, axial hypotonia and stridor (hypomyelinating disorders such as PMD), progressive macrocephaly with hypotonia (Canavan, Alexander, MLC), extreme irritability with stiffness and feeding difficulties (infantile Krabbe), and congenital infection-like encephalopathy with fever, seizures and hepatosplenomegaly (AGS) [9, 31]. In young children, loss of walking or progressive diplegia with neuropathy suggests late infantile MLD, and ataxia worsening after fever or a fall suggests VWM [31, 55]. At school age, declining school performance, new-onset hyperactivity or behavioural changes in a boy should prompt consideration of cerebral X-ALD, which may be mistaken for attention deficit hyperactivity disorder (ADHD) [31, 33]. In adults, spastic paraparesis with polyneuropathy, ataxia with leukoencephalopathy, presenting dysautonomia or a frontal syndrome with leukoencephalopathy should prompt consideration, respectively, of AMN and other inborn errors, LBSL or CTX, ADLD and CSF1R disease [11, 21, 58].

The first clinical step when faced with leukoencephalopathy is to distinguish genetic from non-genetic causes, an important and often difficult task for which expert neuroradiological opinion is decisive [31]. Some genetic diseases mimic environmental causes, such as the “destructive” lesions associated with COL4A1 or RNase T2 deficiency, which mimics congenital cytomegalovirus (CMV) infection [9, 31]. In children, high T2 signal in cerebral WM after 18 months of age is abnormal and warrants investigation [31].

4.2. Historical clues

A targeted history includes a three-generation family history, consanguinity (most leukodystrophies are recessive), affected siblings and ancestry (for example, Canavan in Ashkenazi Jews, POLR3A in French Canadians) [10, 14, 31]. Pregnancy and delivery provide information on congenital infection or perinatal asphyxia; birth weight and head circumference, neonatal history (seizures, fever, coagulopathy, hepatosplenomegaly), the presence and duration of a period of normal development, regression (declining school performance, worsening gait, increasing feeding difficulties), behavioural and psychiatric changes (irritability, aggression, hallucinations), visual or hearing loss and type of motor disturbance (weakness, ataxia, spasticity, dystonia) guide diagnosis [31]. Some clues are almost specific: recurrent vomiting in Alexander disease, adrenal dysfunction in X-ALD, early autonomic dysfunction in ADLD, chronic CSF lymphocytosis or recurrent “aseptic meningitis” in AGS, abrupt loss of milestones after fever or a fall in VWM, congenital cataract in HCC, hypodontia and hypogonadism in POLR3 leukodystrophy, and severe seizures in free sialic acid storage disorders [14].

4.3. Examination clues

Examination seeks macrocephaly (Canavan, Alexander, MLC, gangliosidoses), peripheral nerve involvement (MLD, infantile Krabbe, mitochondrial diseases, connexin 32, HYCCIN deficiency), ocular signs (rotatory nystagmus in PMD, cherry-red spot in GM1/GM2 gangliosidoses, retinal dystrophy, cataract), gait pattern, movement disorders, skin features (photosensitivity in Cockayne syndrome, hyperpigmentation in X-ALD, chilblains in AGS, ichthyosis in Sjögren-Larsson), abnormal dentition (POLR3, oculodentodigital dysplasia), adult palatal myoclonus (Alexander), xanthomas (CTX) and retinal vascular abnormalities (CRMCC) [14, 31]. Formal ophthalmological and cognitive assessments are part of the initial examination [31].

4.4. Warning signs and emergencies

Some situations require immediate action. Acute neurological deterioration after fever, head trauma, surgery or anaesthesia in a patient with VWM, LBSL or another leukodystrophy with decompensations requires admission with high-dependency monitoring, fever control and a search for the cause [3, 42]. Signs of adrenal insufficiency (hyperpigmentation, hyponatraemia, hypoglycaemia, prolonged recovery from anaesthesia) in a boy require measurement of cortisol and very long-chain fatty acids (VLCFA) [9, 33]. Gadolinium enhancement on MRI in a boy with X-ALD indicates active cerebral disease and opens a narrow therapeutic window [59, 60]. Early cerebral X-ALD lesions in a boy with ADHD-like symptoms qualify for transplantation, which is why MRI should not be delayed [33]. Respiratory warning signs also include frequent cough or cough at mealtimes, sialorrhoea, snoring, apnoeas, somnolence, weak cough or pneumonias, prolonged recovery from infections, neck flexor weakness, tachypnoea and stridor [33]. A sudden and persistent change in tone should prompt a search for an intercurrent cause, such as pain, infection or new CNS pathology [33].

4.5. Newborn screening

Newborn screening has become a route of entry for three treatable leukodystrophies. In the United States, the Recommended Uniform Screening Panel (RUSP) added X-ALD in February 2016, infantile Krabbe disease in 2024 and MLD on 16 December 2025; inclusion in the RUSP does not oblige states to screen immediately, a decision taken at state level [54, 61, 62]. Table 4.1 summarises the methods and available results.

Table 4.1. Newborn screening for leukodystrophies

In the Health Resources and Services Administration (HRSA) public consultation, 96 of 98 comments received favoured adding MLD to the RUSP; the advisory committee (ACHDNC) was disbanded after completing the evidence review, so the final recommendation came from HRSA [54]. MLD incidence estimates differ: the FDA states 1/40 000 in the United States, consistent with detection in the German and Austrian pilots, while a 2026 North American review states around 1/100 000 [64, 66, 67]. Whenever leukodystrophy is suspected, the GLIA consensus recommends reassessing the patient's newborn screening result [9].

4.6. LBSL — when to suspect

LBSL should be considered in a child, adolescent or adult with slowly progressive ataxia, predominantly lower limb spasticity and posterior column dysfunction, particularly if initial development was normal and reflexes are preserved [3]. Clinical features alone are non-specific, and MRI, with selective brain stem and spinal cord tract involvement, distinguishes the disease [3]. Childhood-onset axonal neuropathy with pes cavus, particularly if pyramidal signs or leukoencephalopathy subsequently develop, should also prompt inclusion of DARS2 among the genes tested [21, 35]. In infants, the combination of microcephaly, severe epilepsy, profound delay and cerebral atrophy or extensive leukoencephalopathy may represent the severe infantile form, which does not always fulfil MRI criteria [3, 34]. Deterioration after minor head trauma, with reduced consciousness and fever, is a specific warning sign [3].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. MRI as the first-line investigation

Brain MRI is the central investigation and should be reviewed by a neuroradiologist with paediatric expertise before further tests are requested, because investigation directed by the imaging pattern is far more efficient than indiscriminate testing [9, 31]. Minimum sequences are sagittal T1, axial T1, axial T2 and FLAIR; according to suspicion, gadolinium (inflammatory component, cerebral X-ALD), susceptibility sequences (calcifications and haemorrhage, AGS), spectroscopy (lactate in mitochondrial diseases, NAA in Canavan disease) and diffusion (for example, in AARS2 leukoencephalopathy) are added [6, 9]. Computed tomography (CT) is the best modality for calcifications, and quantitative susceptibility mapping is an alternative [6]. Spinal MRI should be performed at least once in all patients with undiagnosed leukodystrophy or gLE, because spinal cord involvement distinguishes several entities, including LBSL [6, 9]. Diffusion tensor imaging and multivariable analyses belong mainly to research [9]. Minimum MRI requirements in leukodystrophy publications and the core protocol for cerebral X-ALD (3D T1, spin-echo T2, 3D FLAIR and diffusion, with contrast in specific settings) have been addressed in recent consensus documents [28, 60].

Interpretation requires knowledge of normal myelination. Most myelination occurs during the first two years of life; on T1, unmyelinated WM is dark and myelinated WM is bright, and on T2 unmyelinated WM has relatively high signal and myelinated WM is dark [31]. T1 shortening precedes T2 shortening during myelination, which is why T1 requires particular attention [9]. A single MRI, especially during the first year of life, cannot distinguish delayed myelination, hypomyelination and early leukodystrophy; serial MRI scans at a minimum interval of 6–12 months are recommended, ideally including at least one after 2 years of age [6, 9]. Once the genetic diagnosis is established, most patients do not require MRI outside clinical trials, with exceptions such as L-2-hydroxyglutaric aciduria (brain tumour risk) and SAMHD1-related AGS (risk of arteriopathy and stroke) [33].

5.2. The three-step pattern algorithm

The 2009 Schiffmann and van der Knaap algorithm, adopted by GLIA in 2015 and adapted by the European ERN-RND network in 2019 (2023 update), organises MRI interpretation around three discriminators [9, 28, 29]. The first is the presence or absence of hypomyelination, namely WM that is hyperintense, isointense or slightly hypointense on T1 relative to the cortex, compared with marked T1 hypointensity in the other conditions; affected WM is T2 hyperintense in all, but less so in hypomyelination [29]. The second, for non-hypomyelinating disorders, is the distinction between confluent and isolated or multifocal abnormalities: multifocal abnormalities suggest acquired (infection, inflammation, vasculopathy) or chromosomal causes, and bilateral symmetrical confluent abnormalities suggest leukodystrophy or gLE [9]. The third is the predominant location of confluent abnormalities, which has high diagnostic value, supplemented by associated findings such as WM rarefaction, cysts, enhancement, calcifications, grey matter involvement and macrocephaly [9, 29].

The general ERN-RND flowchart starts with the question “is this a leukodystrophy?”; if so, a suggestive MRI pattern leads to targeted testing (genetic or biochemical), and the absence of a suggestive pattern or a diagnosis after targeted testing leads to NGS or exome sequencing, with the suggestion to measure arylsulfatase A, β-galactosidase, VLCFA and homocysteine; undiagnosed cases are discussed by a European multidisciplinary panel through the Clinical Patient Management System (CPMS) [28]. Table 5.1 reproduces the pattern classification in the European document, adapted from Schiffmann and van der Knaap; disorders are listed as possibilities (“may be”), rather than certainties.

Table 5.1. MRI patterns of non-hypomyelinating disorders (T2 hyperintensity and clear T1 hypointensity), according to ERN-RND [28, 29]

The 2014 GeneReviews (retired) and GLIA summarised the classical confluent patterns with reference examples: diffuse (MLC, VWM), frontal (Alexander), parieto-occipital (X-ALD), periventricular (MLD), subcortical (Canavan, Kearns-Sayre), temporal (AGS), brainstem (APBD), cerebellum and peduncles (ADLD, middle peduncles) and posterior fossa (peroxisomal disorders) [9, 14]. The posterior predominance of X-ALD and anterior predominance of infantile Alexander disease are among the most recognisable patterns, although X-ALD may have the reverse predominance [31].

5.3. Hypomyelination

In hypomyelinating disorders, the second step examines progression and associated structures: improvement in myelination over time suggests MCT8 deficiency or delayed myelination; severe cortical atrophy suggests primary neuronal disease; persistent hypomyelination without atrophy suggests PMD; cerebellar atrophy suggests POLR3; and basal ganglia atrophy suggests H-ABC [9, 14]. In a study of 112 patients with hypomyelination of known cause, blinded MRI scoring and cluster analysis identified 10 clusters, 8 of which corresponded predominantly to one disorder; discriminating features were early cerebellar atrophy, homogeneous WM T2 signal, abnormal basal ganglia signal, pontine abnormalities and additional T2 lesions in the deep WM [68]. Spectroscopy shows low choline as a marker of hypomyelination [53].

Table 5.2. MRI patterns of selected hypomyelinating disorders

5.4. Associated findings and MRI criteria for specific entities

The third step of the algorithm considers associated findings: rarefaction and cysts visible on FLAIR (VWM), calcifications and haemosiderin visible on CT (AGS), contrast enhancement (X-ALD, Alexander, Krabbe), macrocephaly, cortical lesions (POLG), dentate nucleus involvement (L-2-hydroxyglutaric aciduria, CTX), thinning of the corpus callosum particularly at the genu (SPG11), non-calcified lesions of the basal ganglia (Alexander), brainstem (ADLD) and spinal cord (LBSL and many others) [14, 21]. In adults, the Queen Square review adds anterior temporal pole involvement (CADASIL, CARASIL, CARASAL), restricted-diffusion “dots” in the deep WM (CSF1R disease, AARS2), the “tadpole brainstem” appearance, with medullary and cervical cord atrophy and a preserved pons (Alexander), involvement of the middle cerebellar and cerebral peduncles (CLCN2), periventricular microcalcifications on CT (CSF1R) and extensive calcifications and cysts in Labrune disease [21]. In VWM, the WM undergoes rarefaction and diffuse cystic degeneration, with a meshwork of strands of preserved tissue and CSF-like signal on all sequences, without isolated well-demarcated cysts, whereas in mitochondrial disorders cysts, when present, are usually isolated and well demarcated [11, 31].

Some leukodystrophies have formal MRI criteria. In Alexander disease, a retrospective study of 217 patients with leukoencephalopathy established that radiological diagnosis requires 4 of 5 criteria: extensive WM abnormalities with frontal predominance; a periventricular rim that is T2 hypointense and T1 hyperintense; basal ganglia and thalamic abnormalities (swelling and T2 hyperintensity, or atrophy); brainstem involvement, particularly the medulla and midbrain; and contrast enhancement of structures such as the ventricular lining, periventricular rim, frontal WM, chiasm, fornix, basal ganglia, thalamus, dentate nucleus and brainstem [39]. ADLD has no formal criteria, but MRI, which may precede clinical manifestations by decades, shows symmetrical T2 hyperintensities in the cerebral WM extending from beneath the motor cortex along the corticospinal tracts to the medullary pyramids, with periventricular extension from frontal to parietal, occipital and temporal regions, spared or mildly affected periventricular rims, early involvement of the superior and middle cerebellar peduncles, spinal cord atrophy with hyperintensity throughout the spinal WM and no enhancement [11]. In MLC, the hemispheric WM is abnormal and swollen, with anterior temporal and sometimes frontoparietal subcortical cysts, which are characteristic but not diagnostic [31, 43].

5.5. Patterns suggestive of inborn errors of metabolism

Metabolic disease textbooks organise leukoencephalopathies due to inborn errors of metabolism (IEM) by patterns of location; the tables of the 2022 edition of Saudubray's textbook are summarised here by pattern [53, 58]. With macrocephaly, the associations described are glutaric aciduria type I (with bitemporal atrophy), L-2-hydroxyglutaric aciduria, Canavan, the mucopolysaccharidoses, MLC and Alexander disease [53]. Hypomyelinating IEM include cerebral folate transport defects (FOLR1), untreated galactosaemia, serine synthesis defects, fucosidosis, sialidosis and several tRNA synthetases (KARS1, RARS1, DARS1, EPRS1, AARS1) [53]. In mitochondrial energy defects, early-onset leukoencephalopathy is an increasingly recognised phenotype, associated with mitochondrial aminoacyl-tRNA synthetases (DARS2, EARS2, AARS2, MARS2), lipoic acid metabolism (ISCA2, NFU1, IBA57), complex IV assembly factors (COA7, COA8) and complex I subunits; cavitating patterns, the “tigroid” appearance due to NDUFA2 and a succinate peak on spectroscopy (SDHA, SDHAF1) are clues [47]. In ribose-5-phosphate isomerase deficiency (RPIA, AR), described in 4 male patients since 2004, multifocal and progressive leukoencephalopathy, with neonatal or infantile onset, is associated with raised peaks in the 3.5–4.0 ppm region on spectroscopy (ribitol and D-arabitol); diagnosis is by polyols in urine, plasma or CSF (urinary ribitol and arabitol more than 10 times above normal), with enzymatic confirmation in fibroblasts or lymphoblasts and RPIA sequencing, and there is no specific treatment [69]. In adults, metabolic lesions are usually bilateral and symmetrical, in specific tracts (pyramidal tracts, cerebellar peduncles, U-fibres), and associated polyneuropathy is highly suggestive of an IEM [58]. Spectroscopy is diagnostic in few disorders: cerebral creatine deficiencies (absent peak), Canavan (high NAA peak) and some complex lipid defects (lipid peak in Sjögren-Larsson) [53].

5.6. First-line biochemical panel

The GLIA diagnostic consensus strongly recommends, in all suspected cases except those with an unequivocal MRI pattern, a minimum panel that must not be omitted because it identifies leukodystrophies with time-sensitive treatment: VLCFA, lysosomal enzymes (including galactocerebrosidase and arylsulfatase A) and cholestanol, together with reassessment of the newborn screening result [9]. The LeukoSeq trial and the 2024 review added urinary organic acids, which identify Canavan disease and L-2-hydroxyglutaric aciduria, and retained this panel as first-line testing, in parallel with genomic testing, also to support interpretation of variants of uncertain significance (VUS) [6, 20]. These tests are completed within a few weeks [6, 9]. Table 5.3 summarises what each test detects and its limitations.

Table 5.3. Biochemical and ancillary tests in the investigation of leukodystrophies

5.7. Genetic testing: yield and choice of technique

Evidence on genomic yield in WM disorders comes mainly from North American and Dutch cohorts (Myelin Disorders Bioregistry Project and the Amsterdam database) and the randomised LeukoSeq trial [20, 32, 36]. Table 5.4 summarises the studies, with the population and denominator for each figure.

Table 5.4. Diagnostic yield of genomic testing in WM disorders

Taken together, these data lead to three conclusions. First, panels restricted to classical leukodystrophies have limited utility because these account for a minority in the cohorts, and their cost and depth advantages are smaller than previously thought [6, 36]. Second, trio genome sequencing, when available, is the preferred choice for an MRI pattern not associated with a known disorder or suggestive of mitochondrial disease, because it provides better exome coverage and detects copy number variants (CNV), mitochondrial DNA and, on some platforms, trinucleotide expansions; exome sequencing is the alternative when genome sequencing is unavailable [6, 20]. Third, a panel or single gene is justified only in the presence of pathognomonic features or a known familial variant [6]. LeukoSeq did not compare genome with exome sequencing, had a small sample and non-standardised standard care, and did not assess proband-only genome sequencing, whose yield will be lower because of de novo variants [20]. In the same cohort, MRI pattern recognition before testing was highly accurate for canonical leukodystrophies (correct diagnosis by at least one reviewer in 8/10; 95% CI 44.4–97.5%) and had no accuracy for disorders that were neither canonical nor mitochondrial (0/18), demonstrating the limits of radiological reasoning alone [20].

5.8. Limitations of genetic testing

A negative result does not exclude a genetic leukodystrophy. The exome misses GC-rich regions (for example, a block with about 95% GC in GJC2), deep intronic variants that create pseudoexons (CSF1R), small regulatory RNAs (SNORD118, Labrune disease), CNVs and complex rearrangements, and the most common intronic variant in LBSL, located in intron 2 of DARS2, is easily missed by the exome or excluded by stringent filtering [3, 20, 32]. The LMNB1 duplication and upstream deletions are typically not detected by routine exome or genome sequencing, which is why GeneReviews does not recommend them for ADLD and requires targeted deletion/duplication analysis that includes the upstream region; the same reasoning applies to the PLP1 duplication, present in 60–70% of patients with PMD [11, 21, 38]. Mitochondrial DNA variants may show variable heteroplasmy across tissues and generations, such as NC_012920.1:m.3243A>G in MT-TL1, detected in 42% of reads in a proband and 15% in the mother [32]. Mosaicism may explain late-onset and mild forms in adults [6]. Reference databases have little ancestral diversity, which hampers interpretation in under-represented populations [6]. In persistently unsolved cases, investigational approaches include RNA sequencing (with tools such as FRASER and OUTRIDER), methylation episignatures, proteomics and long-read sequencing, whose application to leukodystrophies is still being studied [6, 41].

5.9. Reporting, VUS, segregation and secondary findings

The report should state the gene, reference transcript with version (preferably MANE Select), HGVS variant nomenclature, zygosity and classification according to ACMG/AMP (American College of Medical Genetics and Genomics/Association for Molecular Pathology) standards, with confirmation by an orthogonal method when indicated and genetic counselling at disclosure [9]. Trio sequencing allows the phase of variants to be established (compound heterozygosity in trans) and de novo variants to be identified [9]. Integration of MRI, phenotype and segregation by a multidisciplinary team allowed cases with missense VUS in highly suggestive genes to be considered resolved in North American cohorts, following clinical, radiological or functional confirmation [20, 36]; this practice does not change the formal variant classification, which should be reviewed when new evidence emerges. First-line biochemistry also provides functional support for VUS (for example, ARSA or galactocerebrosidase activity) [6]. In the 2016 exome cohort, pathogenic or likely pathogenic secondary findings in actionable genes were found in 3/71 families (KCNQ1 and SDHB), in 2.1% of unaffected adults and 3.7% of affected children, which should be discussed during consent [36]. Periodic reanalysis of negative exome or genome data is recommended, given the yield of 10–15% within 1–3 years [6].

5.10. Integrated algorithm

Table 5.5 summarises the current algorithm for children and adults, combining the LeukoSeq algorithm, the 2024 review, the Queen Square adult approach and the European flowchart.

Table 5.5. Integrated diagnostic algorithm

LeukoSeq proposed, as a practical rule, performing first-line biochemistry unless the MRI pattern is confidently recognised, to avoid missing X-ALD, CTX, MLD and Krabbe, using targeted testing when the presentation and MRI are characteristic of a disorder, and proceeding to rapid trio genome sequencing when the pattern does not correspond to a known disorder [20]. This rule updates the 2015 approach, which envisaged broad-spectrum NGS (panel, exome or genome) only after an ambiguous MRI with a clearly genetic aetiology [9]. The divergence from the 2019 adult approach, which still preferred panels for their speed and cost-effectiveness, reflects evolving access to genome sequencing rather than disagreement about the evidence [6, 21].

5.11. LBSL — diagnostic criteria

The clinical diagnosis of LBSL is based on MRI criteria formulated in 2012 and revised in 2021, and requires all major criteria and at least one supportive criterion (Table 5.6) [3].

Table 5.6. MRI criteria for LBSL [3]

The criteria have limitations. In prenatal and early infantile forms, cortical degeneration followed by profound cerebral atrophy or leukoencephalopathy may occur without brainstem abnormalities, and patients may not meet the criteria [3, 34]. Spectroscopy shows elevated lactate in abnormal WM in most patients, but not all, and has low discriminatory value: if MRI meets the criteria, LBSL should be considered with or without lactate; if it does not, elevated lactate suggests a generic mitochondrial leukoencephalopathy [3]. The severe infantile Norwegian case from 2026 had no lactate on spectroscopy [46]. In the 2012 review of approximately 3000 MRI scans, clinical and laboratory findings did not distinguish patients with and without DARS2 variants, whereas MRI did: involvement of the medial lemniscus decussation, anterior spinocerebellar tracts, superior and inferior cerebellar peduncles and globus pallidus was discriminatory [34].

The molecular diagnosis is established by biallelic pathogenic or likely pathogenic variants in DARS2; two VUS or one pathogenic variant with one VUS neither establish nor exclude the diagnosis, and in these cases functional testing of mtAspRS activity in lymphoblasts may provide confirmation [3]. Options are DARS2 sequencing with deletion/duplication analysis if 0–1 variants are found, a multigene panel, or exome/genome sequencing [3]. Because the most common variant is in intron 2, if exome sequencing finds a single heterozygous DARS2 variant in a patient with leukoencephalopathy, intron 2 should be specifically analysed; the Queen Square review recommends single-gene testing when the presentation is suggestive, because this region is poorly covered by NGS [3, 21]. If exome sequencing is not diagnostic, an exome array may detect deletions of one or more exons, such as exon 12 deletion [3]. Sequencing detects approximately 90% of pathogenic variants, a proportion of detectable variants that is not equivalent to clinical sensitivity: in 4 of 43 families, the second variant was not found in genomic DNA, but was detected in cDNA in 2 of 3 of these families analysed, and one patient meeting the MRI criteria had no variants in DNA or cDNA [3]. In the GeneGuidance database, there is no ClinGen actionability assessment for DARS2, either for adults or children, consistent with the absence of a disease-modifying intervention [8].

5.12. Diagnosis of other entities with dedicated guidelines

The 2025 international VWM guideline recommends (strength A) genetic confirmation whenever MRI suggests the disease; considers the diagnosis confirmed with compatible MRI and two pathogenic, likely pathogenic or uncertain-significance variants in an EIF2B1–EIF2B5 gene (strength B); requires additional efforts to find the second variant when only one is identified (A); directs investigation for another cause if MRI is compatible but no variants are found (A), MRI review when variants are found (A) and provision of genetic counselling (A) [42]. In POLR3 leukodystrophy, testing order depends on ancestry: POLR3A first in French Canadians, a founder population, POLR3B first in Europeans, and POLR1C if both are negative, with a panel or exome/genome sequencing as an alternative; POLR3A accounts for 41% of cases, POLR3B for 49% (with multiexon deletions in approximately 3%), POLR1C for 5%, and 5% have no identified gene [10]. In ADLD, targeted analysis identified LMNB1 duplication in 28 of 33 published families and upstream deletion in 5 [11]. In Canavan disease, diagnosis is based on urinary NAA or brain spectroscopy and/or biallelic ASPA variants [41].

5.13. Prenatal and preimplantation diagnosis

Prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible when familial variants have been identified; for some disorders, prenatal diagnosis may also be enzymatic or biochemical (amniotic fluid NAA in Canavan disease), and samples should be agreed with the laboratory beforehand [14, 31]. Without a molecular diagnosis, prenatal diagnosis is not possible [31]. Counselling aspects are developed in section 8.

6. Differential Diagnosis

6.1. Acquired leukoencephalopathies

The first differential diagnosis of a leukodystrophy is acquired leukoencephalopathy, because several are treatable and some reversible [9, 58]. Acute disseminated encephalomyelitis responds to high-dose corticosteroids, and vitamin B12 deficiency is reversible with vitamin replacement [9]. Acquired demyelinating disorders (multiple sclerosis, ADEM, neuromyelitis optica, MOG antibody-associated disease) are distinguished by abrupt onset, a multiphasic course, multifocal lesions and variation or improvement with treatment [6, 14]. Table 6.1 summarises the main acquired differential diagnoses.

Table 6.1. Acquired leukoencephalopathies to exclude

6.2. Hereditary disorders with abnormal WM that are not leukodystrophies

Several genetic diseases alter the WM without being leukodystrophies in the GLIA sense [1, 14]. Hereditary vasculopathies (CADASIL, CARASIL, COL4A1/COL4A2, CRMCC/Coats plus, Fabry disease) produce multifocal lesions that become confluent only in the end stage and are not primarily glial [1, 14]. Axenfeld-Rieger syndrome (FOXC1, PITX2) may be associated with diffuse leukoencephalopathy suggestive of cerebral microangiopathy, described with a FOXC1 variant [70]. Primary neuronal diseases (GM1 and GM2 gangliosidoses, infantile neuronal ceroid lipofuscinoses, LAMA2 congenital muscular dystrophies, Sanfilippo) disturb myelination or cause secondary Wallerian degeneration [1, 14]. MCT8 deficiency presents with markedly delayed myelination that progresses to near completion [14]. IEM with secondary WM involvement, such as untreated organic acidaemias and aminoacidopathies (phenylketonuria, maple syrup urine disease, homocystinuria), urea cycle disorders, Wilson disease and biotinidase deficiency, are gLE, and the WM changes may regress with treatment, which makes their recognition urgent [1, 9]. Diseases with simultaneous grey and white matter involvement, such as MELAS, POLG-related diseases and familial haemophagocytic lymphohistiocytosis, complete this group [14]. In the LeukoSeq cohort, 39% of diagnoses were non-canonical gLE, which shows that this differential is the rule rather than the exception [20].

6.3. Comparative table of leukodystrophies

Table 6.2. Leukodystrophies: genes, inheritance, distinguishing features and discriminating test

6.4. Differential diagnosis of hypomyelinating disorders

Table 6.3. Differential diagnosis of hypomyelinating leukodystrophies [10]

6.5. Differential diagnosis in adults

In adults, the Queen Square review organises genetic leukoencephalopathies into disorders with an available biochemical marker (X-ALD, Krabbe, MLD, CTX, MTHFR deficiency and homocystinuria, mitochondrial DNA disorders) and disorders diagnosed only genetically (CSF1R, CADASIL, CARASIL, CARASAL, TREX1-related retinal vasculopathy with cerebral leukodystrophy, COL4A1, Alexander, ADLD, AARS2, VWM, LBSL, MNGIE, PMD, PMLD, APBD, H-ABC, CLCN2, 4H, Labrune, Gordon Holmes, Nasu-Hakola) [21]. The ADLD GeneReviews provides the most detailed comparison of leukodystrophies presenting in middle age with pyramidal, cerebellar and autonomic signs (Table 6.4).

Table 6.4. Differential diagnosis of adult leukodystrophies with pyramidal and cerebellar signs [11]

In adult spastic paraparesis, polyneuropathy on electromyography, leukoencephalopathy on MRI and an acute or subacute course with sensory ataxia suggest an IEM; dopamine synthesis deficiencies may mimic spastic paraparesis and respond very well to levodopa [58]. Prolonged neonatal jaundice is a retrospective clue to CTX [58].

6.6. LBSL — differential diagnosis

The clinical features of LBSL alone are non-specific and overlap with hereditary ataxias and spastic parapareses; MRI distinguishes the disease [3]. Table 6.5 summarises the differential diagnoses and their discriminating features.

Table 6.5. Differential diagnosis of LBSL

7. Treatments and Risk Management Strategies

7.1. Principles

The GLIA care consensus is based on the principle that, although definitive cures are unavailable for most, all leukodystrophies are treatable in the sense that preventive and symptomatic care changes quality of life and morbidity, and that the absence of a diagnosis should not prevent a comprehensive care plan [33, 71]. The revised 2017 version supersedes the 2015 version and is based mainly on expert consensus and extrapolation from other chronic neurological disorders, such as cerebral palsy and neuromuscular disorders, without formal evidence grades [33]. Disease-modifying therapies are available for a growing number of leukodystrophies, almost all effective only when started before symptoms or at a very early stage, making rapid diagnosis and newborn screening part of treatment [30, 63]. The therapeutic approach therefore has three layers: promptly identifying treatable causes, offering specific therapy when available and organising multidisciplinary supportive care for everyone.

7.2. Treatable causes not to miss

Table 7.1 brings together the leukodystrophies and gLE with specific treatment that the initial investigation should identify.

Table 7.1. Leukodystrophies and leukoencephalopathies with specific treatment

Data on IEM with secondary WM abnormalities come from metabolic textbook tables and the GLIA gLE list; their inclusion here serves as a reminder to investigate them in the initial panel, and treatment of each follows its own recommendations [1, 53]. When primary prevention is possible, such as HSCT in X-ALD, Krabbe disease and MLD, testing should be offered to asymptomatic at-risk relatives [14].

7.3. Approved therapies by disorder: regulatory status

Table 7.2 compares the situation in the United States and European Union, with the practical consequences of differences, as at 9 October 2026.

Table 7.2. Approved disease-modifying therapies in leukodystrophies: USA, EU and practical consequences

For the remaining leukodystrophies, including Canavan, VWM, PMD and PMLD, POLR3, H-ABC, MLC, AGS, CSF1R disease and LBSL, there is no FDA- or EMA-approved disease-modifying therapy in October 2026; investigational treatments are described in section 7.9 [3, 77].

7.4. Zilganersen in Alexander disease

Zilganersen (Zanvastro®, Ionis) is an antisense oligonucleotide targeting GFAP mRNA that reduces production of glial fibrillary acidic protein, whose accumulation causes the disease; it was approved by the FDA on 3 September 2026 for the treatment of Alexander disease in paediatric and adult patients, the first approved therapy for this disease [22]. The recommended dose is 50 mg intrathecally every 3 months, administered by lumbar puncture by experienced professionals, following dilution in the artificial CSF supplied with the vial, as a bolus over 1 to 3 minutes; the final injected volume depends on patient age, according to the label's dilution table [22]. Each vial contains 56 mg in 2.8 mL (20 mg/mL); a missed dose should be administered as soon as possible, with the interval restarted from that date [22]. There are no labelled contraindications [22].

The pivotal trial (study 1, NCT04849741) was randomised, controlled and double-blind, with 49 patients aged 2 to 65 years with Alexander disease confirmed by phenotype, MRI and a pathogenic GFAP variant, and an open-label substudy with 4 patients under 2 years [22]. The 25 mg (n=8) and 50 mg (n=24) cohorts were randomised 2:1 against control (n=17), with administration every 12 weeks for 60 weeks, followed by 60 open-label weeks, a 120-week extension and 28 weeks of follow-up [22]. Median age was 11 years, 65% were female and 86% white, and mean baseline walking speed was 1.2 m/s [22]. In the primary stratum (patients aged 5 years or older with gross motor impairment), the percentage change in walking speed on the 10-metre walk test at week 61 was −2.1% with 50 mg versus −35.4% with control, a difference of 33.3% (95% CI 1.44–65.25; p=0.041), in an analysis of 17 and 13 patients [22]. In the subgroup aged 2 to 4 years, assessed using dimensions D and E of the GMFM-88 scale, there was improvement with zilganersen (n=4) and decline with control (n=3), with a difference of 22.9 points (standard error 5.2) [22]. Use below 2 years is based on the trial from 2 years of age, pharmacokinetic modelling of CSF exposure and substudy safety data [22].

Adverse reactions with a frequency ≥10% and at least 10 percentage points above control (50 mg n=24 versus control n=17) were vomiting (50% versus 29%), back pain (50% versus 18%), cough (38% versus 18%), headache (29% versus 12%), post-lumbar puncture syndrome (29% versus 6%), arthralgia (25% versus 6%) and oropharyngeal pain (21% versus 0%) [22]. The main warning is aseptic meningitis: one patient experienced severe aseptic meningitis during the double-blind period, which recurred during the extension and required interruption and premedication with intravenous dexamethasone before subsequent doses; CSF pleocytosis was observed in 29% of patients treated with 50 mg after the first 2–4 doses, versus 18% in the control group [22]. Compatible symptoms should prompt investigation and treatment according to usual practice [22]. Safety is based on 53 patients, 38 with one year and 18 with two years of exposure (median 60 weeks); there are no human pregnancy data [22]. An expanded access programme (NCT07487389) changed to “approved for marketing” status in September 2026 [22, 77]. Evidence limitations are the small trial size, wide confidence interval and duration of the controlled period (60 weeks), insufficient to assess long-term survival and function.

7.5. Chenodeoxycholic acid and cholic acid

In CTX, chenodeoxycholic acid (CDCA) replaces the deficient bile acid and suppresses cholestanol and bile alcohol synthesis; HMG-CoA reductase inhibitors are mentioned as adjuncts [9]. Approved dosing differs between jurisdictions, as shown in Table 7.3.

Table 7.3. Chenodeoxycholic acid dosing in CTX

In peroxisomal biogenesis disorders, cholic acid (Cholbam®) is approved in the United States as an adjunct at 10–15 mg/kg once daily or in two doses (11–17 mg/kg in familial hypertriglyceridaemia), with liver enzymes, bilirubin and INR monitored monthly for the first 3 months, quarterly for the next 9, every six months for the following 3 years and then annually; safety and efficacy for extrahepatic manifestations, including neurological manifestations, are not established [73].

7.6. Cellular and gene therapies by disorder: overview

Allogeneic HSCT is the reference treatment for early cerebral X-ALD, late-onset forms of MLD and presymptomatic infantile Krabbe disease, and its outcomes depend on the stage at the time of transplantation [9, 62]. Atidarsagene autotemcel is administered as a single intravenous dose of transduced autologous CD34+ cells, after mobilisation, apheresis and myeloablative conditioning, with minimum doses of 4.2×10⁶ (presymptomatic late infantile form), 9×10⁶ (presymptomatic early juvenile form) and 6.6×10⁶ (early juvenile form with early symptoms) CD34+ cells/kg and a maximum of 30×10⁶/kg; warnings include thrombosis and thromboembolism, encephalitis, serious infection, veno-occlusive disease, delayed platelet engraftment and risk of neutrophil engraftment failure, and surveillance for haematological malignancy includes annual full blood count and integration site analysis for at least 15 years [24, 66]. Elivaldogene autotemcel requires a minimum dose of 5.0×10⁶ CD34+ cells/kg with myeloablative and lymphodepleting conditioning, and the label carries a boxed warning for haematological malignancies (myelodysplastic syndrome and acute myeloid leukaemia, potentially fatal) diagnosed between 14 months and 10 years after administration, with a full blood count at least every 3 months and assessment of clonal expansion [59]. Leriglitazone, a PPARγ receptor agonist authorised in the EU, should be stopped at least 5 days before HSCT conditioning and has oedema and weight gain as common adverse effects, with quarterly liver monitoring during the first 6 months [23].

7.7. Supportive care by system (2017 GLIA consensus)

Tone and movement. Spasticity should be assessed by a physiotherapist using scales such as the Hypertonia Assessment Tool (which distinguishes dystonia, spasticity and rigidity), the modified Ashworth scale and modified Tardieu scale, and dystonia using the Global Dystonia Rating Scale or MD-CRS, although none is validated in leukodystrophies [33]. In mild forms without axial hypotonia, oral baclofen or diazepam is used with physiotherapy and daily stretching; botulinum toxin or phenol treats focal spasticity (adductors, for hygiene; calf muscles, for walking); in refractory spasticity, intrathecal baclofen allows higher doses with fewer systemic effects, at the cost of infection risk and mechanical failure with withdrawal syndrome; lengthening surgery, tenotomy or neurotomy is reserved for selected cases, weighing anaesthetic risk [33]. Dystonia is frequently underdiagnosed; trihexyphenidyl is first-line for generalised dystonia (well tolerated, requires treatment of constipation), with levodopa and tetrabenazine as alternatives, and baclofen and benzodiazepines at higher doses than used for spasticity; focal dystonia is treated with botulinum toxin and, in selected cases, intrathecal baclofen or, rarely, deep brain stimulation [33]. Dopamine D2 receptor blockers should be avoided because of tardive dyskinesia risk and, in POLR3 leukodystrophy, worsening extrapyramidal manifestations [10, 33]. Ataxia is treated with rehabilitation, fall prevention and assistive devices [11, 14].

Bone and orthopaedics. The risk of low bone mass and fractures is elevated because of immobility, little sunlight exposure, nutritional deficiencies, corticosteroids, antiepileptics (phenobarbital, phenytoin, carbamazepine, valproate) and prolonged proton pump inhibitor (PPI) use; VWM, AARS2 leukoencephalopathy and POLR3 involve endocrine dysfunction with secondary osteoporosis, and CTX involves granulomatous bone lesions [33]. Elevated parathyroid hormone is the initial marker of low calcium intake or absorption, serum calcium alone is insensitive, and 25-hydroxyvitamin D should be measured at least annually [33]. Bone densitometry (DEXA) of the lumbar spine and total body less head should be performed at approximately 5 years, the first age with reference values, and repeated according to results; with contractures or scoliosis, measurement is at the forearm or lateral distal femur, and plain radiography is unsuitable for screening, except lateral spine radiography for vertebral fractures [33]. In an epidemiological study cited by the consensus, scoliosis occurred in 70% and hip dysplasia in 89%, without the population being specified in the text [33]. Hip surveillance, adapted from cerebral palsy, consists of examination every 6 months and annual radiography from 2 years until skeletal maturity; conservative approaches (adductor release, tone control) are preferred below 5 years, and reconstructive surgery may be considered above 6 years when there is pain, restricted mobility or another risk [33]. The spine should be examined at every visit, with anteroposterior and lateral radiographs if scoliosis is suspected; orthoses optimise posture but do not alter progression, and surgery is generally considered with a Cobb angle above 40–50°, taking into account thoracic insufficiency syndrome risk [33]. Gait is assessed using the GMFCS and 10-metre walk test, and weight-bearing is maintained with orthoses, walking frames and standing frames [33].

Feeding and gastrointestinal care. Feeding risk factors should be assessed at diagnosis and every 3–6 months by a speech and language therapist and occupational therapist, with instrumental studies (videofluoroscopy, modified barium swallow, fibreoptic endoscopic evaluation) if aspiration is suspected [33]. In POLR3 leukodystrophy, dysphagia is highly variable, even within the same day, because of cerebellar incoordination with fatigue [10]. Documented aspiration, difficulty maintaining weight or inefficient oral intake warrants referral for gastrostomy or jejunostomy, which is often delayed and associated with greater risk; gastrostomy may coexist with oral feeding, and there is no evidence supporting concomitant fundoplication in the absence of significant reflux [33]. The gastrostomy decision should consider general condition, expected course and the wishes of the family and patient, and the consensus recommends discussing “quality-of-life care” rather than “end of life”, early and in the outpatient setting [14, 33]. Gastro-oesophageal reflux is diagnosed clinically and treated with positioning, consistency adjustment, antacids, antisecretory agents (ranitidine, lansoprazole, omeprazole) and prokinetics; prolonged PPI use requires monitoring calcium, magnesium, iron and vitamin B12, and failure of medical treatment leads to Nissen fundoplication, usually with gastrostomy [33]. Constipation, defined in practice as two or fewer bowel movements per week, is treated with diet, hydration and fibre, polyethylene glycol or lactulose, sodium phosphate enemas and senna or bisacodyl [33]. In MLD, the gallbladder is involved in more than half of patients, warranting gastroenterological follow-up with scheduled ultrasound; polyps smaller than 5 mm are followed annually and those larger than 5 mm prompt consideration of laparoscopic cholecystectomy; approximately one quarter have acid–base disturbances [33].

Sialorrhoea. First-line measures are oromotor exercises, positioning, review of sialogogue drugs and treatment of constipation, scoliosis and reflux [33]. Anticholinergics are used cautiously because they thicken secretions and cause sedation: oral or transdermal scopolamine, trihexyphenidyl, 1% ophthalmic atropine sublingually and glycopyrrolate, with fewer central effects; in refractory sialorrhoea, botulinum toxin A into the salivary glands and surgery, which in POLR3 may include relocation of the parotid and submandibular ducts with or without sublingual gland excision [10, 33].

Bladder and kidneys. Urinary tract infection may present with fever or hypothermia; urine culture above 100,000 CFU/mL is a reasonable threshold, and samples from catheterised patients should not be used because of colonisation [33]. Two or more infections per year in girls, one or more in boys, or delayed stream, urgency and secondary enuresis warrant urology referral, with renal and bladder ultrasound, voiding studies and urodynamics; retention is treated with catheterisation [33]. In ADLD, neurogenic bladder is treated with a bladder regimen and antispasmodics such as solifenacin, rarely with antibiotic prophylaxis [11]. In AGS, lupus-like glomerulonephritis requires regular proteinuria testing [33].

Respiratory care. Primary respiratory failure occurs late, particularly with peripheral neuropathy (MLD); bulbar dysfunction with central apnoeas characterises type II Alexander disease, and dystonic dysphonia (H-ABC, 4H) and hypotonia predispose to obstructive apnoeas [33]. Swallowing should be reassessed every 3–12 months, more often in high-risk patients [33]. Prevention includes annual influenza vaccination, positioning and hand hygiene, palivizumab in selected cases, chest physiotherapy, oscillatory vest, cough assist and home suction [33]. For the prevention of severe respiratory syncytial virus (RSV) infection, the 2017 consensus mentions only palivizumab. Nirsevimab (Beyfortus®) has been authorised in the EU since 31 October 2022 for neonates and infants in their first RSV season (50 mg below 5 kg and 100 mg from 5 kg, as a single IM dose) and, since 1 August 2024, also for children up to 24 months of age who remain vulnerable in their second season, at 200 mg [78]. Clesrovimab (Enflonsia®), at a fixed dose of 105 mg IM, was approved by the FDA on 9 June 2025 and authorised in the EU on 15 April 2026, for the first season only [79, 80]. Palivizumab (Synagis®), 15 mg/kg monthly during the season, remains indicated for prematurity of 35 weeks or less, bronchopulmonary dysplasia treated within the previous 6 months and haemodynamically significant congenital heart disease, under 2 years of age [81]. Ventilation options (invasive, CPAP, BiPAP, oxygen) should be discussed in advance, in the context of care goals, bearing in mind that, unlike in neuromuscular disorders, they frequently arise in the presence of severe cognitive impairment [33].

Communication, sleep and pain. Speech and language therapy assessment includes dysarthria and apraxia, and augmentative and alternative communication (symbols, voice output, eye-gaze or scanning access) should be assessed whenever communication does not meet daily needs [33]. Sleep disturbances, common in children with severe multisystem disability, include obstructive and central apnoea and the irritability of infantile Krabbe disease and AGS; they are treated with sleep hygiene, a diary and, off-label, clonidine, tricyclics, benzodiazepines or melatonin, as no drugs are FDA-approved for paediatric insomnia [33]. Pain is assessed with age-appropriate scales (CRIES, FLACC, numerical scales, Wong-Baker), hidden causes are sought (dental abscess, constipation, pancreatitis, fractures, dislocation, ulcers, urinary tract infection, neuropathy in Krabbe and MLD), and treatment follows World Health Organization principles; gabapentin is used for neuroirritability and neuropathic pain, and benzodiazepines and neuroleptics for agitation [33]. Peripheral neuropathy and pain should be systematically assessed in Krabbe, MLD, AARS2, polyglucosan body disease and PMD, and the case described in section 10 suggests that neuropathic pain is also underestimated in infantile LBSL [33, 46].

Seizures. There is no evidence for prophylaxis before the first seizure, because most patients do not develop seizures; when a seizure occurs, provoking factors should be sought (fever, electrolyte disturbances, drug withdrawal, infection, particularly urinary), and provoked seizures rarely require daily prophylaxis [33]. Rescue medicines are rectal diazepam and buccal or intranasal midazolam [33]. Mimics to distinguish include Sandifer syndrome, breath-holding spells, syncope, movement disorders, behavioural events, pain episodes, parasomnias and hypoglycaemia [33]. In VWM, the 2025 guideline diverges from this principle and recommends maintenance antiepileptic treatment after even a single seizure (strength A, 83% agreement), because seizures predict a worse course [42].

Dysautonomia. This is prominent in late Alexander disease and ADLD [33]. “Autonomic storms” require investigation for triggers and may be prevented with gabapentin, cyproheptadine, baclofen, β-blockers or clonidine, and treated acutely with diphenhydramine, paracetamol or ibuprofen [33]. Orthostatic hypotension is treated with hydration and salt, compression stockings and, in ADLD, fludrocortisone or a vasopressor; erectile dysfunction with sildenafil; and anhidrosis by avoiding overheating, using cooling vests and antipyretics during infections [11, 33]. Alacrima is treated with lubrication and night-time occlusion [33].

Behaviour. Irritability in infantile Krabbe disease is treated first-line with gabapentin because of its safety profile, despite not having been assessed in trials; unvalidated alternatives are pregabalin, topiramate, tricyclics and valproate, and benzodiazepines are used cautiously [33]. In advanced X-ALD, risperidone and valproate are useful according to clinical experience, and psychiatric symptoms in adult forms require psychiatric care [33].

Endocrine, eyes, teeth and heart. Endocrine surveillance includes the adrenal axis in X-ALD, growth, puberty and thyroid in 4H (testosterone or oestradiol, LH and FSH at the expected age of puberty, by 13 years in girls and 14 in boys; TSH and T4), ovarian function in VWM, AARS2, mitochondrial diseases and 18q–, thyroid in CTX and AGS and, after transplantation, growth, thyroid, gonadal function and osteoporosis [33]. AGS is associated with autoimmunity (thyroid, lupus, coeliac disease, inflammatory bowel disease), with annual TSH [33]. Specific ophthalmological problems (severe myopia and retinal detachment in 4H, glaucoma in AGS, cataracts in HCC and CTX, nystagmus in PMD) require ophthalmological follow-up, with lubrication if blinking is reduced; dental problems require a dentist experienced in special needs [33]. Regular cardiology is recommended in AGS, 18q–, infantile sialic acid storage disease, fucosidosis and mitochondrial diseases, with attention to accelerated coronary disease in CTX [33].

Care coordination. The consensus recommends a biopsychosocial assessment by a social worker, a written longitudinal care plan, a list of “red flags” for contacting the specialist centre, participation in research and natural history registries, transition planning (inpatient and outpatient, paediatric and adult care, home and institutional care) and early palliative care, which is distinct from terminal care [33]. The team includes genetics and genetic counselling, neurology, complex care paediatrics, respiratory medicine, gastroenterology, speech and language therapy, endocrinology, physical and rehabilitation medicine, orthopaedics, physiotherapy and occupational therapy, palliative care and social services [33]. School and work require advocacy, augmentative communication and adaptations [14].

7.8. Symptomatic drugs: dose information in the sources

Consensus sources for leukodystrophies specify drugs but, with few exceptions, do not give doses, and doses of symptomatic drugs (baclofen, diazepam, tizanidine, trihexyphenidyl, glycopyrrolate, fludrocortisone, solifenacin, sildenafil) should follow national and paediatric formularies, which the sources consulted do not replace [11, 33, 38]. Table 7.4 compiles the doses actually present in the sources consulted.

Table 7.4. Doses described in consensus sources for supportive care

The vitamin D thresholds used by the consensus are deficiency below 20 ng/mL (50 nmol/L), insufficiency between 21 and 29 ng/mL (52.5–72.5 nmol/L) and sufficiency between 30 and 100 ng/mL [33]. These values reproduce the 2011 Endocrine Society guideline, cited by the 2017 consensus; the table does not define a maximum dose for any age group and, in infants, indicates repeating the measurement at 4 weeks [33]. According to the same consensus, glycopyrronium is approved in children over 3 years of age [33].

7.9. Prevention of decompensation and agents to avoid

Episodic worsening after mild head trauma and infections has been suggested in several leukodystrophies, but is consistently documented only in VWM; as a precaution, avoidance of these triggers is recommended in the others [14]. In LBSL, deteriorations after head trauma have been described, with unknown exact risk and no guidelines on contact sports, so counselling and decisions should be individual [3]. In PMD and ADLD, fever transiently worsens symptoms, and in ADLD avoidance of exposure to infections is recommended [11, 38]. In MLC, avoidance of contact sports is recommended [43]. In POLR3-related leukodystrophy, D2 blockers (neuroleptics such as haloperidol and risperidone; antiemetics such as metoclopramide) and foods with a choking risk should be avoided [10].

The 2025 international VWM guideline, developed through a Delphi process with a multidisciplinary panel and published by ERN-RND in July 2026, contains 43 recommendations graded as strength A (strongly recommended), B and C (suggested); those on prevention and acute management are summarised in Table 7.5 and provide a model for other leukodystrophies involving decompensations [42].

Table 7.5. Recommendations for prevention and management of VWM (2025 guideline) [42]

The 2019 VWM GeneReviews recommended overlapping measures: vaccines, low-dose maintenance antibiotics during winter, antibiotics for minor infections, antipyretics, helmets outdoors in children and close neurological monitoring for several days during febrile infections and after trauma or surgery with anaesthesia [55].

7.10. Pregnancy, perioperative care and transition

Pregnancy data in leukodystrophies are scarce in the sources consulted. The VWM guideline recommends close surveillance during pregnancy and delivery at a centre with obstetrics, neurology and intensive care [42]. The zilganersen label has no human pregnancy data, and the leriglitazone label has no data below 2 years and contraindicates use in heart failure [22, 23]. Perioperatively, prolonged recovery from anaesthesia may be the first sign of adrenal insufficiency in X-ALD and peroxisomal biogenesis disorders, and patients with X-ALD should have a corticosteroid stress-dose letter [9, 33]. In VWM, major surgery and anaesthesia are recognised triggers, and postoperative monitoring is mandatory [42]. In MLD, acid–base disturbances worsen under stress [33]. Transition from paediatric to adult care should be planned with the written care plan and a team familiar with the disease [33].

7.11. Trials supporting therapies and negative trials

Table 7.6. Trials and treatment studies in leukodystrophies

These studies do not allow indirect comparisons with one another, because they differ in disease, population, outcome and design. In metachromatic leukodystrophy and X-ALD, comparisons are with natural history cohorts rather than randomised controls; in the leriglitazone trial, the upper confidence limit of 100% reflects the small sample, and comparison is with an assumed rate of 10%, not a control group [23, 66]. For zilganersen, the relative difference of 33 percentage points in walking speed should be interpreted in light of the absolute decline of approximately one third in the control group in just over a year and an analysis involving 30 patients [22]. Negative or inconclusive studies (amino acids in LBSL, curcumin in PMD, lithium and triacetate in Canavan) demonstrate that mechanistic plausibility is insufficient and that absence of effect in small samples does not exclude a stabilising effect, as acknowledged by the amino acid pilot authors [82].

7.12. Developments in the last 12 months and research, by evidence level

Authorisations. Zilganersen for Alexander disease, approved by the FDA on 3 September 2026, with no application identified at the EMA [22]. Leriglitazone for cerebral X-ALD in the EU (boys aged 2–12 years with non-enhancing lesions and a neurological function score of 0–1), with a positive CHMP opinion on 23 July 2026 and authorisation on 21 September 2026, under exceptional circumstances [23].

Screening policy decision. Addition of MLD to the US RUSP on 16 December 2025 [54].

Regulatory designation without authorisation. Fosigotifator (ABBV-CLS-7262, eIF2B activator, Calico/AbbVie) received FDA Breakthrough Therapy designation for VWM, according to a company announcement of 23 June 2026 that was not confirmed in an FDA source; it is in an open-label phase 1b/2 trial (NCT05757141; adults, children and infants; 50 planned participants; recruiting; primary completion expected in November 2027), without released numerical clinical data; the expanded access programme (NCT06594016) is no longer available and there have been two single-patient treatments in Cree leukoencephalopathy/VWM [83, 84].

Ongoing trials without published results. In Canavan disease, BBP-812 (AAV9-ASPA; phase 1/2 CANaspire trial, NCT04998396, 26 participants, recruiting, primary completion expected in October 2026), with sustained reductions in urinary, CSF and brain NAA and improved myelination reported by the company, and oligodendrocyte-targeted rAAV-Olig001-ASPA (phase 1/2, NCT04833907, 24 participants, by invitation), with NAA reduction and improved myelination presented at a conference; these are preliminary data and announcements, not authorisation, and no marketing application was identified for BBP-812 [41, 77]. In PMD, ION356, an intrathecal anti-PLP1 antisense oligonucleotide, is in phase 1 (NCT06150716, 24 participants, until 2028), with an associated biomarker study (NCT05659901); deferiprone is in a trial based on models [38, 77]. In Krabbe disease, the PBKR03 gene therapy trial (NCT04771416) is suspended and FBX-101 (NCT05739643) is active, not recruiting [77]. Intrathecal umbilical cord-derived oligodendrocyte cells (DUOC-01), combined with cord blood transplantation, are in phase 1 (NCT02254863) [77]. In adult cerebral X-ALD, the phase 3 CALYX leriglitazone trial (NCT05819866) has results expected in 2028 [77].

Preclinical research in LBSL. AAV9-DARS2 gene therapy increased gene expression, improved mitochondrial function and axonal growth and reduced lactate release in patient-derived neurons, and improved behaviour and cortical neurodegeneration in mice with neuronal Dars2 deletion six months after a single intracerebroventricular injection [27]. Antisense oligonucleotides modulating DARS2 expression showed in vitro efficacy and in vivo brain biodistribution, according to GeneReviews [3]. Cantharidin was identified in a fluorescent reporter drug screen as the most potent modulator of DARS2 splicing, as a proof of principle [19]. None of these approaches has reached a human trial.

Natural history registries. Registries are active for LBSL (NCT03624374, Kennedy Krieger, 100 participants, until 2029; NCT05750979, Amsterdam) and the Myelin Disorders Biorepository Project (NCT03047369) [77]. The absence of identified marketing applications for BBP-812, fosigotifator and leriglitazone in the United States results from searching public databases and does not prove their non-existence [77].

7.13. LBSL — management

There is no disease-modifying treatment for LBSL [3]. The initial evaluation recommended by GeneReviews includes neurological examination, brain and spinal MRI with spectroscopy of abnormal WM if not already performed, EEG if seizures are suspected, assessment by physical and rehabilitation medicine, physiotherapy and occupational therapy (weakness, spasticity, gait), speech and language therapy if dysarthria is present, cognitive assessment if impairment is suspected, genetic counselling and family support [3]. Wearable sensors (accelerometers) have been validated for remote gait and coordination assessment, correlating with ataxia scales [3]. Treatment is symptomatic: physiotherapy and rehabilitation to improve function and prevent contractures and scoliosis, positioning and mobility aids, antiepileptics for seizures (usually easy to control), speech and language therapy for dysarthria, early intervention and an individualised education plan when intellectual disability is present, social services, care coordination and ongoing assessment of palliative care and home nursing needs [3]. Because muscle biopsies, fibroblasts and lymphoblasts do not show mitochondrial dysfunction, there is no rationale for a vitamin and cofactor “mitochondrial cocktail” [3]. Head injury should be avoided, with an individual decision on contact sports [3]. In severe infantile disease, neuropathic pain may be intense and underestimated [46]. Aspartate supplementation did not demonstrate efficacy in the 2025 pilot, and gene and antisense therapies are preclinical [27, 82].

7.14. Guidelines and divergences

Table 7.7. Applicable guidelines and consensuses and points of divergence

8. Reproductive Options and Family Counselling

8.1. Recurrence risks according to inheritance mechanism

With an established molecular diagnosis, recurrence risk follows the inheritance mechanism [31]. In autosomal recessive disorders, which account for the majority, parents are obligate heterozygotes, and each sibling of a patient has a 25% probability of being affected, 50% of being an asymptomatic carrier and 25% of being a non-carrier; a patient's children are obligate carriers, and aunts and uncles have a 50% probability of being carriers [3, 10]. In X-linked disorders (X-ALD, PMD), mothers may be heterozygous and heterozygous women may have their own manifestations, such as middle-age myelopathy in X-ALD or mild to moderate disease in PMD [31, 38]. In autosomal dominant disorders, the risk to an affected individual's offspring is 50%; when the variant arises de novo (Alexander, TUBB4A), sibling risk is low, allowing for germline mosaicism [39, 44]. In ADLD, all probands whose parents were tested inherited the alteration from an affected parent, no de novo cases have been described, and if the alteration is not detected in the parents the sibling risk is approximately 1% (theoretical germline mosaicism) [11].

8.2. Counselling without a molecular diagnosis

When a leukodystrophy remains undiagnosed, the Oxford Desk Reference first recommends a recent expert MRI review and panel, exome or genome sequencing [31]. If no cause is found even then, counselling is based on the probability of a genetic aetiology: with parental consanguinity, autosomal recessive inheritance is likely and recurrence risk is probably 25%; with a previously affected sibling, the risk is also probably 25%, recessive inheritance being more likely than X-linked inheritance or germline mosaicism; without either factor, environmental factors or de novo variants should be considered, but counselling should explain that risk may reach 25% [31]. Without a diagnosis, carrier detection is possible only through the pedigree and prenatal diagnosis is not possible [31]. Periodic reassessment may yield a diagnosis through phenotype evolution, repeat MRI, data reanalysis or new genes, and DNA should be banked from probands without a molecular diagnosis [3, 6, 31].

8.3. Carrier testing, cascade screening and population screening

Carrier testing requires prior identification of familial variants [3, 31]. Partner screening is problematic except in populations with common variants, and enzyme testing rarely provides sufficient discrimination for partners [31]. The classical example of population screening is Canavan disease in Ashkenazi Jews, in whom p.(Glu285Ala) and p.(Tyr231Ter) allow targeted carrier screening, with a carrier frequency between 1/40 and 1/82; however, most new diagnoses now occur in people of other ancestries, with diverse and private variants [40, 41]. When primary prevention is possible (HSCT in X-ALD, Krabbe and MLD), testing should be offered to asymptomatic at-risk relatives [14]. In ADLD, MRI may be abnormal years before clinical manifestations, information that should be communicated to at-risk relatives [11].

8.4. Predictive testing and minors

In adult-onset disorders without benefit from early treatment, such as ADLD, predictive testing is possible with formal counselling but should not be performed in asymptomatic minors [11]. In disorders with presymptomatic treatment (X-ALD, MLD, Krabbe), testing of minor siblings is justified by direct benefit, and newborn screening has made it routine in several programmes [14, 64]. X-ALD newborn screening also identifies forms that may never develop cerebral disease, requiring counselling about uncertainty and a surveillance plan [61]. Affected or carrier young adults should receive genetic counselling, ideally before pregnancy [3].

8.5. Prenatal diagnosis and preimplantation genetic testing

With familial variants identified, prenatal diagnosis (chorionic villus sampling, amniocentesis) and preimplantation genetic testing are possible; for some disorders, prenatal diagnosis may be based on the enzyme or a metabolite, with samples agreed with the laboratory [3, 10, 31]. Intrafamilial variability in severity should be discussed, because a positive result does not accurately predict disease severity [3, 10].

8.6. LBSL — counselling

LBSL is autosomal recessive; both parents should be tested [3]. If only one parent is a carrier, a de novo variant in the proband, postzygotic parental mosaicism or uniparental isodisomy should be considered [3]. Heterozygous carriers are asymptomatic and are not at risk of developing the disease [3]. Siblings have a 25% probability of being affected, 50% of being carriers and 25% of being non-carriers, with variable expression: in the Sudanese family described, siblings with the same variants ranged from almost asymptomatic to disabled [3, 46]. The Finnish carrier frequency of 1:95 should not be used to estimate prevalence, because almost all patients are compound heterozygotes and no Finnish homozygote has been found [3, 8]. Discovery of allelic axonal CMT implies that relatives with isolated peripheral neuropathy may have biallelic DARS2 genotypes and benefit from assessment [35]. Prenatal diagnosis and preimplantation testing are possible with known variants [3].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Group frequency: what each study measures

There is no validated worldwide prevalence of leukodystrophies, and published figures measure different things: minimum incidence in surveys, regional hospital incidence, lifetime risk in national surveillance or predicted incidence from allele frequencies [9, 14]. Table 9.1 summarises the main studies.

Table 9.1. Group epidemiological studies

North American consensuses cite a combined incidence range of 1/50 000 to 1/7663, corresponding to the extremes of the German and Utah studies [9, 33]. The difference between estimates of 1/50 000 and 1/4733 mainly reflects what is counted (disorders with biochemical markers versus all genetic leukoencephalopathies, observed versus predicted) rather than true geographical variation [15, 18]. Hospital costs associated with leukodystrophies in the United States were estimated at over 59 million dollars annually [33].

9.2. Frequency of specific entities

Table 9.2. Frequency of selected leukodystrophies

9.3. Founder populations and consanguinity

Several leukodystrophies have founder variants with implications for testing. In LBSL, carrier frequency in Finland is 1:95, and c.492+2T>C and c.455G>T are linked to shared haplotypes common in patients from north-eastern Europe [3]. In Canavan disease, the Ashkenazi variants p.(Glu285Ala) (80–85% of alleles) and p.(Tyr231Ter) contrast with p.(Ala305Glu), more common in non-Ashkenazi Europeans, and with the intronic insertion of approximately 2600 bp, which may be the most common pathogenic variant across several ancestries [31, 41]. In POLR3 leukodystrophy, French Canadians are a founder population for POLR3A, determining testing order [10]. In MLC, MLC1 founder variants have been described in the Agrawal community in India, Libyan and Turkish Jews, Koreans and Egyptians [43]. In the United Kingdom, the number of children with leukodystrophy was relatively high in consanguineous Pakistani families [17]. LBSL is an exception among rare recessive disorders, because only one consanguineous family has been described and only six individuals from four families were homozygous [3].

9.4. Portugal, Brazil and gaps

In the sources consulted, data from Portugal and Brazil are scarce. The LBSL GeneReviews cites a series of three Brazilian patients published in 2007, whose text was not directly consulted [3]. The study of 66 patients with LBSL included collaborators from Portuguese centres, without national data in the abstract [19]. The list of Member States in the ERN-RND flowchart document (64 centres and 4 partners in 24 Member States) does not include Portugal [28]. The lack of data in the sources consulted does not demonstrate an absence of published data. Genomic reference databases have little ancestral diversity, complicating variant interpretation in under-represented populations, including those of African and Latin American ancestry [6].

9.5. Special populations

Adults. Diagnostic yield is lower, the differential diagnosis with acquired causes broader and the preferential genes different (CSF1R, LMNB1 duplication, AARS2, NOTCH3), as described in sections 3.4 and 6.5 [6, 21]. Women. Ovarian failure (ovarioleukodystrophy) in VWM and AARS2 requires endocrine assessment, and the VWM guideline recommends close surveillance during pregnancy [9, 42]. Heterozygous women with X-ALD and PMD may have their own manifestations [31, 38]. Children with severe disability. Communication, sleep, pain and feeding require structured assessment, and ventilation should be discussed considering cognitive impairment, as described in section 7.7 [33]. Screened newborns. Newborn screening programmes create a new population of presymptomatic children or children with predicted late-onset forms, requiring structured surveillance and support for family decision-making [62, 64].

9.6. Reference networks and patient organisations

In Europe, the relevant reference network is ERN-RND (European Reference Network for Rare Neurological Diseases), which has a leukodystrophy group and provides diagnostic flowcharts, standards of care in several languages, minimum MRI requirements, international recommendations on X-ALD, the GMFC-MLD scale, the VWM management guideline, a standard operating procedure for nerve conduction studies in leukodystrophies, recommendations on MLD newborn screening and treatment eligibility and neuroradiology panels through CPMS [28]. Hereditary metabolic disorders with leukoencephalopathy are also within the scope of MetabERN [6]. Patient organisations include the European Leukodystrophy Association (ELA; France, with branches in Belgium, Luxembourg, Switzerland, Spain and Germany), Alex TLC (United Kingdom), the Global Leukodystrophy Initiative (theglia.org), the United Leukodystrophy Foundation (ulf.org), Cure LBSL (curelbsl.org), the Yaya Foundation (POLR3 leukodystrophy), the ADLD Center, Leukodystrophy Australia and the Leukodystrophy Foundation of Canada; the Myelin Disorders Bioregistry Project is a research registry and biobank [3, 10, 11, 28, 33]. In the United Kingdom, the Oxford Desk Reference listed Climb and Contact a Family in 2017 [31].

10. Illustrative Clinical Case

Heier and colleagues described a Norwegian case of severe infantile LBSL in 2026, published open access with MRI and RNA analysis figures [46]. The patient was the first child of healthy, non-consanguineous Norwegian parents. At 3 months of age, he was assessed for excessive weight gain of approximately 500 g per week, intermittent stridor and hypotonia; analysis of breast milk showed 6.9 g of fat and 98 kcal per 100 g, approximately 50% above usual [46]. Serial laryngoscopies documented progressive vocal cord paresis, and at 6 months the child had stridor, obesity (weight Z-score 4.25), areflexia and marked hypotonia, without head control [46].

Brain MRI showed diffusion restriction and T2 hyperintensity in the centrum semiovale and periventricular WM, posterior limbs of the internal capsules, central tegmental tracts and inferior cerebellar peduncles (https://pmc.ncbi.nlm.nih.gov/articles/PMC13299155/figure/F1/), and spectroscopy at echo times of 35 and 144 ms showed no lactate [46]. Nerve conduction studies elicited no motor response, particularly in the lower limbs, electromyography showed extensive spontaneous activity and EEG was normal [46]. At 6 and a half months, the child cried constantly, had complete vocal cord paralysis and spinal MRI showed dorsal column hyperintensity along almost the entire cord, without enhancement (https://pmc.ncbi.nlm.nih.gov/articles/PMC13299155/figure/F2/); tracheostomy was performed with palliative intent [46].

At 7 months, trio exome sequencing, analysed with a 3670-gene panel, identified compound heterozygosity in DARS2: c.259G>A p.(Asp87Asn) and c.294G>A p.(Glu98=), a synonymous variant at the last nucleotide of exon 3 (described on NM_018122.4; same nomenclature on NM_018122.5) [46, 48]. RT-PCR on parental blood showed that c.259G>A caused partial exon 3 skipping (r.228_294del, p.Arg76SerfsTer6), probably through disruption of an exonic splicing enhancer, and that c.294G>A caused complete exon 3 skipping (https://pmc.ncbi.nlm.nih.gov/articles/PMC13299155/figure/F3/) [46]. Neither variant is the common intronic variant, consistent with the observation that patients with severe infantile disease do not carry it [3, 46].

Neuropathic pain was the dominant clinical problem. Paracetamol, ibuprofen, midazolam, morphine and chloral hydrate were insufficient; continuous subcutaneous morphine was increased to 14 mg/kg/day during the final 2 months, and low-dose oral ketamine had no effect [46]. After an intravenous ketamine bolus, ketamine via nasogastric tube in divided doses produced a rapid response, with less discomfort and improved heart rate and oxygenation, and was progressively increased to doses that the authors state exceeded twice the maximum daily dose in the reference paediatric palliative formulary [46]. These doses relate to an individual palliative care case and are not a recommendation. The child was discharged at 8 months with a home care team and a do-not-resuscitate decision, and died at 9 and a half months from respiratory failure, without an autopsy [46].

The case illustrates several points of this article. Infantile LBSL may lack lactate on spectroscopy and may present with severe peripheral involvement, departing from the classical presentation and approaching the CNS–peripheral continuum described in allelic CMT [35, 46]. Vocal cord paresis had not previously been described in LBSL, and pain may be underestimated in the disease [46]. The c.259G>A variant had been described in only one other case with early onset and rapid progression, reinforcing the association between more severe loss-of-function variants and infantile disease [3, 46]. Finally, functional confirmation of the effects of two variants that, from their sequence, would appear respectively missense and synonymous demonstrates the value of RNA analysis in classifying splicing variants [46].

11. Prognosis and Long-Term Follow-up

11.1. Group prognosis

Prognosis depends on the disorder, form and age at onset, and is highly variable within each entity [9, 30]. In the Utah cohort (1999–2007), one third of children died, with mean age at death of 8.2 years, almost half had epilepsy and a gastrostomy, and half remained undiagnosed; these figures describe an era before exome sequencing and gene therapies and include gLE [16]. In the multicentre cohort of 704 children (2013–2020), 29% had seizures, particularly in astrocytopathies and at older ages [52]. In disorders with approved therapy, prognosis depends on treatment timing: in late infantile MLD treated presymptomatically with gene therapy, all children were alive at 6 years, versus 58% in natural history [66]. Adult forms are generally slowly progressive, with survival over decades in ADLD and adult-onset LBSL [3, 11]. In disorders with decompensations (VWM), onset before 4 years, acute declines and seizures predict a worse course [42]. In POLR3 leukodystrophy, early-onset forms carry a risk of death in young adulthood, and late forms may survive into the fourth or fifth decade [10]. In Canavan disease, most patients with the typical form survive to 10 years and a minority to adulthood [41].

11.2. LBSL prognosis

There are no prospective natural history data for LBSL [3]. Childhood-onset disease progresses slowly, with partial or complete wheelchair dependence in adolescence, the twenties or later, complete dependence being rare and generally occurring in adulthood; adult-onset disease causes mild impairment, and no patients are known to have become wheelchair-dependent; severe infantile disease is often fatal in the first years, with death at 20 months in the earliest-onset patient in the 2012 series and at 9 and a half months in the 2026 case [3, 19, 34, 46]. Intellectual ability is normal in most, and epilepsy is uncommon and easy to control in non-infantile forms [3]. Recovery after post-traumatic episodes may be only partial [3].

11.3. Surveillance

Table 11.1 compiles the surveillance intervals recommended in the sources, for the group and for entities with dedicated guidance. Unvalidated intervals are identified as such.

Table 11.1. Surveillance in leukodystrophies

After genetic diagnosis, most patients do not need routine MRI outside clinical trials [33]. Surveillance of the adrenal axis in X-ALD, MRI surveillance of presymptomatic forms and follow-up after gene therapy have their own protocols; the 2017 consensus proposed ACTH and morning cortisol every 6 months, an interval that differs from the 2022 international recommendation [33].

11.4. Palliative care and planning

Palliative care should be introduced early, alongside other care, and is distinct from terminal care; discussion of gastrostomy, ventilation and care goals should take place in the outpatient setting, before a crisis [33]. The written care plan, shared between the family, specialist centre and local care providers, should include the “red flags” warranting urgent contact and instructions for foreseeable decompensations, such as fever in VWM or adrenal crisis in X-ALD [33, 42].

12. Summary and Final Considerations

Leukodystrophies are genetic disorders primarily affecting CNS WM; the 2015 GLIA definition restricts them to disorders with primary glial cell or myelin abnormalities and distinguishes them from genetic leukoencephalopathies, whereas the 2017 definition includes all primary genetic WM disorders and proposes a cell-type classification (myelin, astrocyte, axon, microglia, vessels). Their collective frequency ranges from 1/50 000 to 1/7663 in observational studies, depending on what is counted, and reaches 1/4733 in genomic estimates, which include variants of uncertain penetrance.

Diagnosis is based on expert-reviewed MRI, interpreted in three steps (hypomyelination or not; confluent or multifocal; location and associated findings), with spinal MRI in undiagnosed cases and serial MRI when the first scan is performed during the first year of life. The minimum biochemical panel (VLCFA, lysosomal enzymes, cholestanol and urinary organic acids) should be performed promptly because it identifies disorders with time-sensitive treatment. For a non-specific MRI pattern, trio genome sequencing is the preferred choice; the LeukoSeq trial showed an overall yield of 76.5% in less than four months, and most diagnoses fall outside classical leukodystrophy genes, limiting restricted panels. Technical limitations (introns, GC-rich regions, CNV such as LMNB1 and PLP1 duplication, mitochondrial DNA, mosaicism) and periodic reanalysis are part of interpreting any negative result.

Treatment combines three layers. Approved therapies exist for cerebral X-ALD, MLD, CTX and, since September 2026 in the United States, Alexander disease, with the antisense oligonucleotide zilganersen; leriglitazone was authorised in the EU in September 2026 for non-enhancing paediatric cerebral X-ALD. Newborn screening for X-ALD, Krabbe disease and MLD transforms these diseases into treatable presymptomatic conditions. For all leukodystrophies, supportive care organised by system, prevention of decompensation and early palliative planning alter morbidity and quality of life.

LBSL, caused by biallelic DARS2 variants, illustrates the group's logic. Diagnosis is radiological, with formal criteria centred on brainstem and spinal cord tracts, and molecular, with attention to the common intronic variant in intron 2, which exome sequencing may miss. The disorder is a leukoaxonopathy due to partial loss of function of a mitochondrial aminoacyl-tRNA synthetase, with a normal respiratory chain in accessible tissues, a spectrum from fatal infantile to mild adult disease and, since 2025, an allelic axonal CMT. There is no disease-modifying treatment; management is symptomatic, with prevention of head injury and no rationale for mitochondrial supplements, and AAV9-DARS2 gene therapy remains preclinical.

13. References

1. Vanderver A, Prust M, Tonduti D, Mochel F, Hussey HM, Helman G, et al. Case definition and classification of leukodystrophies and leukoencephalopathies. Mol Genet Metab. 2015;114(4):494-500. doi:10.1016/j.ymgme.2015.01.006. PMID: 25649058. PMC4390457.
2. van der Knaap MS, Bugiani M. Leukodystrophies: a proposed classification system based on pathological changes and pathogenetic mechanisms. Acta Neuropathol. 2017;134(3):351-382. doi:10.1007/s00401-017-1739-1. PMID: 28638987. PMC5563342.
3. Engelen M, Abbink TEM, Salomons GS, van der Knaap MS. Leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2010 [updated 2021 Feb 18; reviewed 2026 Aug 20]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK43417/
4. Scheper GC, van der Klok T, van Andel RJ, van Berkel CG, Sissler M, Smet J, et al. Mitochondrial aspartyl-tRNA synthetase deficiency causes leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation. Nat Genet. 2007;39(4):534-9. doi:10.1038/ng2013. PMID: 17384640.
5. van der Knaap MS, van der Voorn P, Barkhof F, Van Coster R, Krägeloh-Mann I, Feigenbaum A, et al. A new leukoencephalopathy with brainstem and spinal cord involvement and high lactate. Ann Neurol. 2003;53(2):252-8. doi:10.1002/ana.10456. PMID: 12557294.
6. Helman G, Orthmann-Murphy JL, Vanderver A. Approaches to diagnosis for individuals with a suspected inherited white matter disorder. Handb Clin Neurol. 2024;204:21-35. doi:10.1016/B978-0-323-99209-1.00009-0. PMID: 39322380. PMC12301652.
7. National Center for Biotechnology Information. DARS2 aspartyl-tRNA synthetase 2, mitochondrial [Homo sapiens] (Gene ID: 55157). NCBI Gene. Bethesda (MD): National Library of Medicine; [accessed 10 Oct 2026]. Available at: https://www.ncbi.nlm.nih.gov/gene/55157
8. GeneGuidance. Gene and condition curation database (version of 2026-09-26): DARS2, POLR3A, POLR3B, POLR1C, LMNB1, GFAP, ASPA, PLP1, EIF2B5, MLC1, TUBB4A.
9. Parikh S, Bernard G, Leventer RJ, van der Knaap MS, van Hove J, Pizzino A, et al. A clinical approach to the diagnosis of patients with leukodystrophies and genetic leukoencephelopathies. Mol Genet Metab. 2015;114(4):501-515. doi:10.1016/j.ymgme.2014.12.434. PMID: 25655951. PMC4390485.
10. Bernard G, Vanderver A. POLR3-related leukodystrophy. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2012 [updated 2017 May 11]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK99167/
11. Raininko R, Gosky M, Padiath QS. LMNB1-related autosomal dominant leukodystrophy. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2016 [updated 2021 Jul 15]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK338165/
12. Orphanet. Leukodystrophy (ORPHA:68356) [group]. Available at: https://www.orpha.net/en/disease/detail/68356
13. Orphanet. Leukoencephalopathy with brain stem and spinal cord involvement-high lactate syndrome (ORPHA:137898). Available at: https://www.orpha.net/en/disease/detail/137898
14. Leukodystrophy overview [chapter retired on 2020 Jan 30; published 2014 Feb 6]. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle.
15. Heim P, Claussen M, Hoffmann B, Conzelmann E, Gärtner J, Harzer K, et al. Leukodystrophy incidence in Germany. Am J Med Genet. 1997;71(4):475-8. PMID: 9286459.
16. Bonkowsky JL, Nelson C, Kingston JL, Filloux FM, Mundorff MB, Srivastava R. The burden of inherited leukodystrophies in children. Neurology. 2010;75(8):718-25. doi:10.1212/WNL.0b013e3181eee46b. PMID: 20660364. PMC2931652.
17. Stellitano LA, Winstone AM, van der Knaap MS, Verity CM. Leukodystrophies and genetic leukoencephalopathies in childhood: a national epidemiological study. Dev Med Child Neurol. 2016;58(7):680-9. doi:10.1111/dmcn.13027. PMID: 26866636.
18. Soderholm HE, Chapin AB, Bayrak-Toydemir P, Bonkowsky JL. Elevated Leukodystrophy Incidence Predicted From Genomics Databases. Pediatr Neurol. 2020;111:66-69. doi:10.1016/j.pediatrneurol.2020.06.005. PMID: 32951664. PMC7506144.
19. van Berge L, Hamilton EM, Linnankivi T, Uziel G, Steenweg ME, Isohanni P, et al. Leukoencephalopathy with brainstem and spinal cord involvement and lactate elevation: clinical and genetic characterization and target for therapy. Brain. 2014;137(Pt 4):1019-29. doi:10.1093/brain/awu026. PMID: 24566671.
20. Vanderver A, Bernard G, Helman G, Sherbini O, Boeck R, Cohn J, et al. Randomized Clinical Trial of First-Line Genome Sequencing in Pediatric White Matter Disorders. Ann Neurol. 2020;88(2):264-273. doi:10.1002/ana.25757. PMID: 32342562. PMC8061316.
21. Lynch DS, Wade C, Paiva ARB, John N, Kinsella JA, Merwick Á, et al. Practical approach to the diagnosis of adult-onset leukodystrophies: an updated guide in the genomic era. J Neurol Neurosurg Psychiatry. 2019;90(5):543-554. doi:10.1136/jnnp-2018-319481. PMID: 30467211. PMC6581077.
22. Ionis Pharmaceuticals. ZANVASTRO (zilganersen) injection, for intrathecal use. Prescribing information [Internet]. DailyMed set id af6ca330-46d3-4ed6-b2af-bd6db3133fd1, published 2026-09-15; Drugs@FDA NDA 220210 (original approval 2026-09-03). Available at: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=af6ca330-46d3-4ed6-b2af-bd6db3133fd1
23. European Medicines Agency. Nezglyal (leriglitazone): EPAR — product information (Summary of Product Characteristics) [Internet]. Amsterdam: EMA; CHMP opinion 2026-07-23; authorisation 2026-09-21 (exceptional circumstances). Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/nezglyal
24. Orchard Therapeutics. LENMELDY (atidarsagene autotemcel) suspension for intravenous infusion. Prescribing information [Internet]. DailyMed set id b3e307a7-561b-43b6-a784-9cb1dcfc5196, published 2026-10-05. Available at: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=b3e307a7-561b-43b6-a784-9cb1dcfc5196
25. Mirum Pharmaceuticals. CTEXLI (chenodiol) tablets. Prescribing information [Internet]. Effective version 2025-12-16; Drugs@FDA NDA 219488 (approval 2025-02-21).
26. European Medicines Agency. Chenodeoxycholic acid Leadiant: EPAR — product information [Internet]. Authorisation 2017-04-10 (exceptional circumstances). Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/chenodeoxycholic-acid-leadiant
27. Garofolo I, Lindsay B, Liang Y, Ratajczak B, Janowski M, Walczak P, et al. AAV9-DARS2 Gene Therapy Rescues Phenotype in Leukoencephalopathy with Brainstem and Spinal Cord Involvement and Lactate Elevation Patient Cells and Neuronal Dars2 Deficient Mice. Ann Neurol. 2026;99(1):59-72. doi:10.1002/ana.78037. PMID: 40948401. PMC13533035.
28. European Reference Network for Rare Neurological Diseases (ERN-RND), Disease group Leukodystrophies (Sevin C, Wolf N, coord.). Diagnostic flowcharts for leukodystrophies [Internet]. Published 2019-02-11; updated 2023-05-10. Available at: https://www.ern-rnd.eu/wp-content/uploads/2019/10/ERN-RND-Diagnostic-Flowchart-Leukodystrophie_FINAL-1.pdf ; page: https://www.ern-rnd.eu/disease-knowledge-hub/leukodystrophies/
29. Schiffmann R, van der Knaap MS. Invited article: an MRI-based approach to the diagnosis of white matter disorders. Neurology. 2009;72(8):750-9. doi:10.1212/01.wnl.0000343049.00540.c8. PMID: 19237705. PMC2677542.
30. van der Knaap MS, Schiffmann R, Mochel F, Wolf NI. Diagnosis, prognosis, and treatment of leukodystrophies. Lancet Neurol. 2019;18(10):962-972. doi:10.1016/S1474-4422(19)30143-7. PMID: 31307818.
31. Firth HV, Hurst JA. Leukodystrophy/leukoencephalopathy. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 208–210.
32. Helman G, Lajoie BR, Crawford J, Takanohashi A, Walkiewicz M, Dolzhenko E, et al. Genome sequencing in persistently unsolved white matter disorders. Ann Clin Transl Neurol. 2020;7(1):144-152. doi:10.1002/acn3.50957. PMID: 31912665. PMC6952322.
33. Adang LA, Sherbini O, Ball L, Bloom M, Darbari A, Amartino H, et al. Revised consensus statement on the preventive and symptomatic care of patients with leukodystrophies. Mol Genet Metab. 2017;122(1-2):18-32. doi:10.1016/j.ymgme.2017.08.006. PMID: 28863857. PMC8018711.
34. Steenweg ME, van Berge L, van Berkel CG, de Coo IF, Temple IK, Brockmann K, et al. Early-onset LBSL: how severe does it get? Neuropediatrics. 2012;43(6):332-8. doi:10.1055/s-0032-1329395. PMID: 23065766.
35. Estévez-Arias B, Sarv S, Bonello-Palot N, Carrera-García L, Ortez C, Expósito-Escudero J, et al. Biallelic Variants in the DARS2 Gene as a Novel Cause of Axonal Charcot-Marie-Tooth Disease. Ann Neurol. 2025;98(6):1335-1351. doi:10.1002/ana.78005. PMID: 40814755. PMC12682944.
36. Vanderver A, Simons C, Helman G, Crawford J, Wolf NI, Bernard G, et al. Whole exome sequencing in patients with white matter abnormalities. Ann Neurol. 2016;79(6):1031-1037. doi:10.1002/ana.24650. PMID: 27159321. PMC5354169.
37. Disorders of nucleic acid metabolism, tRNA metabolism and ribosomal biogenesis. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 39, section 39.3.2.
38. Wolf NI, van Spaendonk RML, Hobson GM. PLP1-related disorders. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1999 [updated 2025 Jun 12]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1182/
39. Srivastava S, Waldman A, Naidu S. Alexander disease. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2002 [updated 2020 Nov 12]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1172/
40. 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, section 22.10. doi:10.1007/978-3-662-63123-2_22
41. Nagy A, Bley AE, Eichler F. Canavan disease. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1999 [updated 2023 Dec 21; reviewed 2026 Jun 23]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1234/
42. van Voorst RJ, Schoenmakers DH, Bonkowsky JL, et al. Consensus-Based Expert Recommendations for Diagnosis and Clinical Management of Vanishing White Matter. Neurology. 2025;105(11):e214320. doi:10.1212/WNL.0000000000214320.
43. Min R, Abbink TEM, van der Knaap MS. Megalencephalic leukoencephalopathy with subcortical cysts. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2003 [updated 2023 Jul 27]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1535/
44. Gavazzi F, Pizzino A, Bally JF, Fraser J, Simons C, Wolf NI, van der Knaap M, Vanderver A. TUBB4A-related neurologic disorders. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2016 [updated 2026 Mar 5]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK395611/
45. Disorders of cellular trafficking. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 44, section 44.2.5.
46. Heier M, Wirgenes KV, Kulseth MA, Chawla MS, Akre H, Vatsgar TT, et al. Vocal Cord Paresis and Neuropathic Pain in Infantile Leukoencephalopathy with Brainstem and Spinal Cord Involvement and Lactate Elevation: A Case Report. Case Rep Neurol. 2026;18(1):241-248. doi:10.1159/000552009. PMID: 42367688. PMC13299155.
47. Disorders of oxidative phosphorylation. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 10, sections 10.1.1, 10.1.2, 10.3.2, Tables 10.3, 10.4.
48. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348 (online tool https://variantvalidator.org, GRCh38, accessed 10 Oct 2026)
49. National Center for Biotechnology Information. ClinVar: VCV000001057 — NM_018122.5(DARS2):c.228-21_228-20delinsC. Bethesda (MD): National Library of Medicine; [accessed 10 Oct 2026]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/variation/1057/
50. National Center for Biotechnology Information. ClinVar: VCV004688052 — NM_018122.5(DARS2):c.1006C>T (p.Arg336Cys); VCV004688053 — NM_018122.5(DARS2):c.713C>T (p.Ser238Phe); VCV004688055 — NM_018122.5(DARS2):c.1508C>T (p.Pro503Leu). Bethesda (MD): National Library of Medicine; [accessed 1 Nov 2026]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/variation/4688053/
51. Happ HC, Christensen GB, Knight S, Novoa A, Isakson D, Nadauld LD, et al. Characterization of Leukodystrophy Penetrance Using Large-Scale Integration of Genomic Population Screening and Electronic Health Records. medRxiv. 2026:2026.07.04.26356970. doi:10.64898/2026.07.04.26356970. PMID: 42465930. PMC13370487.
52. Hackett AN, Yan H, Liu K, Corre C, Gao K, Grzyb C, et al. Risk of Seizures and Epilepsy in the Leukodystrophies. Pediatr Neurol. 2026;174:20-28. doi:10.1016/j.pediatrneurol.2025.09.021. PMID: 41166783.
53. 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, 1.5.2.4, 1.5.2.6, 1.5.6, Tables 1.15, 1.20, 1.25, 1.30.
54. Health Resources and Services Administration (HHS). Addition of Metachromatic Leukodystrophy to the Recommended Uniform Screening Panel. Fed Regist. 2025 Dec 22;90(243):59847-8. FR Doc. 2025-23574. Available at: https://www.federalregister.gov/documents/2025/12/22/2025-23574/addition-of-metachromatic-leukodystrophy-to-the-recommended-uniform-screening-panel
55. van der Knaap MS, Fogli A, Boespflug-Tanguy O, Abbink TEM, Schiffmann R. Childhood ataxia with central nervous system hypomyelination / vanishing white matter. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2003 [updated 2019 Apr 4]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1258/
56. Wolf NI, Biancheri R, Zara F, Bruno C, Gazzerro E, Rossi A, van der Knaap MS, Minetti C. Hypomyelination and congenital cataract. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2008 [updated 2021 Jan 14]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK2587/
57. Min R, Depienne C, Sedel F, Abbink TEM, van der Knaap MS. CLCN2-related leukoencephalopathy. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2015 [updated 2021 May 20]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK326661/
58. Mochel F, et al. Inborn errors of metabolism in adults: a diagnostic approach to neurological and psychiatric presentations. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 2, sections 2.2.3, 2.3.5, 2.3.8, Table 2.7.
59. bluebird bio. SKYSONA (elivaldogene autotemcel) suspension for intravenous infusion. Prescribing information [Internet]. Effective version 2026-07-21 (openFDA drug label); CBER page https://www.fda.gov/vaccines-blood-biologics/skysona (approval letters of 2022-09-16 and 2024-04-05 listed).
60. Yska HAF, Golse M, Galanaud D, Amartino HM, Bergner C, Bruschi F, et al. Use of Brain MRI in Cerebral Adrenoleukodystrophy: International Recommendations for Screening, Monitoring, and Research. Neurology. 2026;106(5):e214657. doi:10.1212/WNL.0000000000214657. PMID: 41650360. PMC12893433.
61. Kemper AR, Brosco J, Comeau AM, Green NS, Grosse SD, Jones E, et al. Newborn screening for X-linked adrenoleukodystrophy: evidence summary and advisory committee recommendation. Genet Med. 2017;19(1):121-126. doi:10.1038/gim.2016.68. PMID: 27337030. PMC5182180.
62. Ream MA, Lam WKK, Grosse SD, Ojodu J, Jones E, Prosser LA, et al. Evidence and Recommendation for Infantile Krabbe Disease Newborn Screening. Pediatrics. 2025;155(4):e2024069152. doi:10.1542/peds.2024-069152. PMID: 40074005. PMC12010248.
63. Laugwitz L, Shenker A, Sluys EF, Pintat S, Whiteman D, Chanson C. Newborn Screening for Metachromatic Leukodystrophy: A Systematic Literature Review. Int J Neonatal Screen. 2025;11(4):103. doi:10.3390/ijns11040103. PMID: 41283365. PMC12641716.
64. Laugwitz L, Martin P, Janzen N, Hegert S, Neiße T, Wacker M, et al. Clinical management of metachromatic leukodystrophy cases identified in the German and Austrian Newborn screening pilots. Orphanet J Rare Dis. 2026;21(1):345. doi:10.1186/s13023-026-04620-y. PMID: 42834411. PMC13640125.
65. Vasco A, Meta A, Berardo C, Camerlengo D, Fumagalli F, Tonduti D, et al. Preliminary Data of the First Year of Newborn Screening for Metachromatic Leukodystrophy (MLD) in Lombardy. Int J Neonatal Screen. 2026;12(3):49. doi:10.3390/ijns12030049. PMID: 42496454. PMC13398140.
66. U.S. Food and Drug Administration. FDA Approves First Gene Therapy for Children with Metachromatic Leukodystrophy [Internet]. Silver Spring (MD): FDA; 2024 Mar 18. Available at: https://www.fda.gov/news-events/press-announcements/fda-approves-first-gene-therapy-children-metachromatic-leukodystrophy
67. Lam WKK, Ream MA, Grosse SD, Ojodu J, Jones E, Smith HS, et al. Evidence Regarding Metachromatic Leukodystrophy Newborn Screening. Pediatrics. 2026;158(Suppl 2):e2026077357E. doi:10.1542/peds.2026-077357E. PMID: 42674587.
68. Steenweg ME, Vanderver A, Blaser S, Bizzi A, de Koning TJ, Mancini GM, et al. Magnetic resonance imaging pattern recognition in hypomyelinating disorders. Brain. 2010;133(10):2971-82. doi:10.1093/brain/awq257. PMID: 20881161. PMC3589901.
69. Wamelink MMC, Valayannopoulos V, Garavaglia B. Disorders of glycolysis and the pentose phosphate pathway. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 7, section 7.9. doi:10.1007/978-3-662-63123-2_7
70. Jones KL, Jones MC, del Campo M. Axenfeld-Rieger syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 834–7.
71. Van Haren K, Bonkowsky JL, Bernard G, Murphy JL, Pizzino A, Helman G, et al. Consensus statement on preventive and symptomatic care of leukodystrophy patients. Mol Genet Metab. 2015;114(4):516-26. doi:10.1016/j.ymgme.2014.12.433. PMID: 25577286.
72. European Medicines Agency. Libmeldy (atidarsagene autotemcel): EPAR [Internet]. Authorisation 2020-12-17. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/libmeldy
73. Mirum Pharmaceuticals. CHOLBAM (cholic acid) capsules. Prescribing information [Internet]. Effective version 2026-01-26; Drugs@FDA NDA 205750 (approval 2015-03-17).
74. European Medicines Agency. Skysona (elivaldogene autotemcel): EPAR — withdrawn [Internet]. Authorisation 2021-07-16; withdrawal 2021-11-18. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/skysona
75. European Medicines Agency. Kolbam (cholic acid): EPAR — withdrawn [Internet]. Authorisation 2015-11-20; withdrawal 2020-03-27. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/kolbam
76. European Medicines Agency. Orphacol (cholic acid): EPAR [Internet]. Authorisation 2013-09-12. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/orphacol
77. ClinicalTrials.gov. Records NCT04849741, NCT07487389 (zilganersen); NCT04998396, NCT04833907 (Canavan); NCT06150716, NCT05659901 (PMD/ION356); NCT04771416, NCT05739643, NCT06308718 (Krabbe); NCT02254863 (DUOC-01); NCT05819866 (leriglitazone CALYX); NCT03624374, NCT05750979 (LBSL); NCT03047369; NCT02699190 [Internet]. Available at: https://clinicaltrials.gov
78. European Medicines Agency. Beyfortus (nirsevimab) — EPAR, product information (Annex I, Summary of Product Characteristics). First authorisation 31/10/2022; extension to the second season (II/0005) 01/08/2024. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/beyfortus [accessed 2026-11-07].
79. U.S. Food and Drug Administration. ENFLONSIA (clesrovimab) injection: prescribing information. BLA 761432; approval 2025 Jun 9 [accessed 7 Nov 2026]. Available at: https://www.accessdata.fda.gov/drugsatfda_docs/label/2025/761432s000lbledt.pdf
80. European Medicines Agency. Enflonsia (clesrovimab): EPAR — product information. EMEA/H/C/006497; CHMP opinion 18/09/2025; authorisation 15/04/2026 [accessed 7 Nov 2026]. Available at: https://www.ema.europa.eu/en/medicines/human/EPAR/enflonsia
81. European Medicines Agency. Synagis (palivizumab) — EPAR and Summary of Product Characteristics (sections 4.1 and 4.2); first authorisation on 13 Aug 1999. EPAR Synagis. Accessed 7 Nov 2026.
82. Catania A, Marchet S, Einvag K, Lamantea E, Salsano E, Ghezzi D, et al. Amino acid supplementation in patients affected by leukoencephalopathies associated with mitochondrial aminoacyl tRNA synthetase pathogenic variants: Pilot clinical trial in adults and review of literature. Mol Genet Metab. 2025;146(3):109259. doi:10.1016/j.ymgme.2025.109259. PMID: 41075682.
83. Calico Life Sciences. Calico Life Sciences announces fosigotifator granted Breakthrough Therapy Designation by the U.S. FDA for the treatment of vanishing white matter disease [Internet]. South San Francisco: Calico; 2026 Jun 23. Available at: https://www.calicolabs.com/press/calico-life-sciences-announces-fosigotifator-granted-breakthrough-therapy-designation-by-the-u-s-fda-for-the-treatment-of-vanishing-white-matter-disease/
84. ClinicalTrials.gov. An Open-Label Exploratory Study of Fosigotifator in Participants With Vanishing White Matter Disease (NCT05757141) [Internet]. Updated 2026-09-24. Available at: https://clinicaltrials.gov/study/NCT05757141
85. Orphanet. Orphadata: epidemiological data of rare diseases (prevalence, incidence and prevalence class by entity). Paris: INSERM; [accessed 10 Oct 2026]. Available at: https://www.orphadata.com/epidemiology/