Krabbe Disease (Globoid Cell Leukodystrophy)

GALC-related galactocerebrosidase deficiency, with reference to saposin A deficiency (PSAP) as a biochemical phenocopy.

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

Krabbe disease is a lysosomal leukodystrophy in which myelin injury results not from diffuse, visible storage of the main substrate, but from a minor metabolite, psychosine, which kills the cells that produce and maintain myelin. This feature explains three central aspects: the disease simultaneously affects the central and peripheral nervous systems, the rate of progression parallels the physiological rate of myelination in the first year of life, and the only treatment that affects the course, haematopoietic stem cell transplantation, substantially changes the prognosis of the infantile form only when performed before symptoms appear [2, 5, 6].

The original description is attributed to the Danish clinician Knud Krabbe, in 1916 [15]. According to the Atlas of Inherited Metabolic Diseases, Krabbe reported five patients, four of whom belonged to two sibling pairs; all were normal at birth and developed rapidly progressive neurological deterioration beginning between 4 and 6 months, leading to death between 1.5 and 2 years. Alongside the clinical presentation, he documented the neuropathological finding that would give the disease its name: infiltration of white matter by large, multinucleated globoid cells [1]. Chemical analysis subsequently showed cerebroside accumulation in these cells, and intracerebral injection of galactocerebroside in animals reproduced globoid cell formation, directing the search for the defect towards the catabolism of this lipid [1].

In 1970, K. Suzuki and Y. Suzuki demonstrated profound galactocerebroside β-galactosidase deficiency in the brain, liver and spleen of three patients, with normal activity in other brain diseases with similar white matter destruction, without a tissue inhibitor or a shift in optimum pH; they concluded that this deficiency explained the disease better than the previous hypothesis of sulphotransferase deficiency [16]. GALC complementary DNA was subsequently cloned and the gene mapped to the long arm of chromosome 14; a deletion of approximately 30 kb was also identified which, according to the Atlas, accounted for a large number of patients from northern Europe, the USA and Mexico [1]. Studies in humans and the twitcher mouse showed increased psychosine in the brain and peripheral nerves, a compound virtually absent from the normal brain and also degraded by the deficient enzyme, supporting the so-called “psychosine hypothesis” [1, 5].

The therapeutic history has two phases. Until the beginning of this century, the Atlas recorded that no effective specific treatment existed and that bone marrow transplantation in patients with later onset had uncertain results [1]. In 2005, Escolar and colleagues showed that umbilical cord blood transplantation in 11 asymptomatic newborns diagnosed through family history altered the natural history, whereas the same procedure in 14 already symptomatic infants produced no substantial neurological improvement [17]. This result led to newborn screening in New York in 2006 and, following two rejections by the US federal advisory committee (2010 and 2023), to inclusion of the infantile form in the RUSP in 2024 [8, 9]. GeneReviews, fully revised on 8 September 2026, and expert recommendations published in 2026 describe the current position: diagnosis must be made within days, the treatment decision is taken with the family within a few weeks, and transplantation changes survival and function but does not cure the disease [2, 6].

2. Molecular Genetics and Pathophysiology

2.1. Galactocerebrosidase and substrates

Galactosylceramide is the characteristic cerebroside of myelin and the central nervous system: a monohexosylceramide in which galactose is attached by a glycosidic bond to carbon 1 of ceramide. Under normal conditions it is degraded in the lysosome to ceramide and galactose by galactosylceramide β-galactosidase (EC 3.2.1.46) [1]. The enzyme, encoded by GALC, also catabolises lactosylceramide and galactosylsphingosine, that is, psychosine; in vivo, galactosylceramide degradation also requires the activator saposin A [5].

The gene occupies 14q31.3 and, according to GeneReviews, GALC is synthesised as an 80 kDa precursor with six potential glycosylation sites, proteolytically cleaved into 50 kDa and 30 kDa subunits that are not active individually and aggregate into a hydrophobic complex of very high molecular mass [2]. The Atlas states that the complementary DNA contains 3795 base pairs and encodes 669 amino acids [1]. The gene has 17 exons, and the 30 kb deletion begins in intron 10 and extends beyond the end of the gene [2, 10].

2.2. From enzyme deficiency to demyelination

In patients, enzyme activity measured in the brain, liver, spleen and kidney corresponds to approximately 5–10% of normal in the historical studies cited by the Atlas, and GeneReviews reports 0–5% of normal in leucocytes of all symptomatic patients [1, 2]. The deficiency causes galactosylceramide accumulation in globoid cells — multinucleated macrophages, 20 to 50 µm in diameter and with up to 20 nuclei, containing tubular crystalloid inclusions — characteristic of white matter demyelinating lesions [1, 5]. The disease is an unusual lipidosis because the main substrate accumulates only in globoid cells and storage cannot be demonstrated in lysosomes of other human tissues; in the twitcher mouse, by contrast, inclusions and accumulation also occur in the kidney and lymphocytes [1].

The decisive cellular injury is attributed to psychosine, which differs from cerebroside by the absence of the fatty acid. It is a highly toxic natural detergent, essentially absent from the normal brain and increased in the brain and peripheral nerves of patients and affected mice [1]. According to Saudubray, recent studies showed that psychosine forms through slow deacylation of galactosylceramide and accumulates in oligodendrocytes and Schwann cells; its apoptotic effect explains the early destruction of oligodendrocytes characteristic of the infantile form [5]. Death of these cells causes central and peripheral nervous system demyelination, astrocytic gliosis and globoid cell infiltration; peripheral nerves become thickened, with endoneurial fibrosis and loss or thinning of myelin sheaths [1, 4].

The relationship between psychosine and phenotype has direct clinical implications: the rate of neurodegeneration parallels the severity of GALC deficiency and psychosine elevation, and later-onset forms have milder biochemical abnormalities and slower progression [6]. In a murine model, haematopoietic transplantation appears to act mainly by restoring phagocytic function rather than through cross-correction of myelinating cells, according to Saudubray [5]. The 2026 recommendations describe a mixed mechanism: enzyme replacement through donor-derived microglia and macrophages, reduced neuroinflammation and lower donor cell uptake in the peripheral nervous system, explaining the more limited effect of transplantation on neuropathy [6].

2.3. Pathogenic variants and genotype–phenotype correlation

Saudubray mentions more than 250 known GALC variants; the evidence review for the RUSP reports more than 1250 recorded variants, of which more than 300 are classified as pathogenic or likely pathogenic [5, 9]. The infantile form results from severe loss of activity, through homozygosity for the 30 kb deletion or compound heterozygosity for that deletion and another severe variant (mainly nonsense or frameshift, but also some missense variants), or through two severe variants [2]. The most frequent allele combines the large deletion with the c.550C>T polymorphism (historical designation 502C>T) and, according to Saudubray, has never been found in Japanese patients [5]. At the time of the GeneReviews revision, all patients identified through newborn screening had at least one copy of this deletion [2].

Table 2.1 lists the variants mentioned in the main sources, with HGVS nomenclature confirmed using VariantValidator on the MANE Select transcript NM_000153.4 and the historical designation where used in the literature. Historical numbering counted amino acids from the signal peptide with a difference of 16 residues, so older references do not coincide with current nomenclature [2, 3].

Phenotype prediction is limited. Even for p.Gly286Asp, associated with milder presentations, GeneReviews emphasises that the individual clinical course cannot be predicted, and patients' genotypes are often complex, combining benign and pathogenic variants [2]. In Japan, the Atlas reports that 12del3ins, p.I66M and I289V, found only in Japanese individuals, accounted for 37% of mutant alleles and, together with p.G270D and p.T652P, 57% of variants in 17 Japanese patients; these designations are historical and were not normalised in this review [1]. Intrafamilial variability is real: siblings with later-onset disease may develop symptoms at very different ages [2].

2.4. Pseudodeficiency and modifying alleles

Three common variants, c.550C>T p.(Arg184Cys), c.742G>A p.(Asp248Asn) and c.1685T>C p.(Ile562Thr), reduce GALC activity measured in vitro without causing disease, and are classified as benign in GeneReviews; their presence complicates interpretation of low activity [2, 3]. The 2026 recommendations define these pseudodeficiency alleles as variants present in more than 1% of the normal population that lower measured activity without causing disease on their own [6]. However, p.Ile562Thr was shown to be causal when on the same chromosome as another mild variant, such as p.Thr112Ala [2]. The Atlas describes later-onset patients with a polymorphism that slightly reduces activity on one allele and a causal variant on the other, and a mutant allele found in a healthy nurse with leucocyte activity below 10% of control, which could lead to diagnostic error, particularly prenatally [1]. The case reported by Schecter et al. in 2026, homozygous for an in cis haplotype with p.Asp248Asn, p.Tyr319Cys and p.Ile562Thr, illustrates how pseudodeficiency and modifying alleles may together approach a risk allele; the authors propose reclassifying p.Tyr319Cys as likely pathogenic, which represents the authors' opinion rather than formal reclassification [19].

2.5. Saposin A deficiency (PSAP)

PSAP encodes prosaposin, a precursor processed in the lysosome into four homologous activator proteins. Saposin A is a cofactor for galactosylceramide and lactosylceramide degradation, and its deficiency produces a variant of Krabbe disease, with four known cases according to Saudubray, all infantile and attributed to variants in the saposin A domain [5]. In the first published case, a six-month-old girl with progressive encephalopathy and abnormal myelination was initially diagnosed with Krabbe disease because of leucocyte GALC deficiency; activity was normal in cultured fibroblasts, loading with labelled lipid indicated abnormal galactosylceramide storage, and a three-base-pair deletion removing a conserved valine at position 11 of saposin A was found [21]. GeneReviews notes that saposin A deficiency may be associated with low GALC and elevated psychosine, and should therefore be excluded when psychosine is high and GALC testing is not diagnostic [2]. Combined prosaposin deficiency is a distinct neurovisceral disease, with nine published cases who died between 4 and 30 weeks of life, hepatosplenomegaly, hypotonia, massive myoclonus, dystonia and seizures; GALC may also be low in leucocytes and fibroblasts, and concomitant elevation of plasma lysoGb3 and lysoGb1 should prompt complete PSAP sequencing [5].

2.6. Heterozygosity and allelic diseases

GeneReviews recognises no other phenotypes associated with germline GALC variants. Heterozygosity has been suggested to be associated with an increased risk of neurodegenerative diseases, open-angle glaucoma and pulmonary artery dilatation, but no formal, statistically robust studies in patients' parents exist; carriers are considered asymptomatic and not at risk of developing Krabbe disease [2].

3. Clinical Spectrum and Manifestations by System

3.1. Classifications by age at onset

The disease forms a continuum, and subtype boundaries vary between sources. This divergence is not merely nominal: the same child with onset at 10 months may be classified as “late-infantile” in an older series and “infantile” in the current classification. Table 3.1 compares the definitions so that each study's figures can be read using the classification from which they arose [6, 15].

The prospective series by Bascou et al. and the transplantation study by Yoon et al. showed that children with onset between 6 and 12 months follow a course resembling the infantile form, whereas those with onset between 13 and 36 months have a different trajectory; reclassification of the late-infantile form as onset between 12 and 36 months is therefore advocated [15, 22].

3.2. Infantile form

The infant appears normal at birth and during the first months. GeneReviews describes four stages. In stage I, the child begins to cry without apparent cause, often attributed to colic or milk allergy; many keep their hands tightly clenched beyond three months; feeding difficulties and gastro-oesophageal reflux may cause weight loss; in some infants, peripheral neuropathy and hypotonia are the first manifestation, lasting several months [2]. Stage II involves rapid neurological deterioration, with decorticate posturing (marked hypertonia, extended and crossed lower limbs, flexed upper limbs and opisthotonus), areflexia, staring episodes, muscle spasms, optic atrophy and sluggish pupillary responses. Stage III adds poor temperature and heart rate control, blindness, deafness and seizures, and stage IV is characterised by profound hypotonia and absence of voluntary movement [2]. The Atlas and Orphanet describe an equivalent sequence in three stages, with hypersensitivity to sound, light and touch triggering screaming and rigidity, recurrent unexplained fever and progression to a decerebrate state around 9–12 months [1, 4].

Manifestations are exclusively neurological, with no visceromegaly or skeletal abnormalities [1, 2]. Head circumference may be large, normal or small; microcephaly is frequent, but macrocephaly and hydrocephalus with intracranial hypertension have been reported [1, 2]. The Atlas additionally reports ichthyosis in one patient, prominent ears, enlarged optic nerves and thickened cauda equina roots on magnetic resonance imaging (MRI), a cherry-red spot reported between 13 and 17 months and palatal myoclonus [1]. In an infant diagnosed prenatally, examination was normal in the neonatal period, tendon reflexes disappeared at five weeks, nerve conduction velocity was abnormal at seven weeks and neck muscle weakness developed at three months; these findings suggest that detailed examination may reveal disease before the reported age at onset [1, 2]. Some newborns with positive screening and confirmed infantile disease had lower limb clonus, feeding difficulties, abnormal nerve conduction, elevated cerebrospinal fluid (CSF) protein or abnormal MRI in the first weeks [2].

Mean age at death without treatment is 24 months, ranging from 8 months (infections and respiratory failure) to 9 years (in a vegetative state); death generally results from aspiration pneumonia or respiratory complications, and frequent vomiting contributes to malnutrition [1, 2]. The evidence review for the RUSP reports median survival of 1.5 years in infants with onset in the first six months [9].

3.3. Later-onset forms

Children with onset between 12 months and 3 years may be normal until they develop gait abnormalities, hemiplegia or diplegia, visual impairment, febrile seizures or tremor; because myelination is very rapid in the first year, symptoms progress faster at that age, and after two years progression is generally slower unless infection occurs [2]. Between 2 and 4 years, presentation may begin with loss of acquired skills and vision, gait abnormalities, tremor and seizures, with esotropia, ocular oscillations and sluggish pupillary responses causing photophobia; upper eyelid paralysis reduces blinking and predisposes to corneal ulcers [2]. In children with onset between 9 months and 4 years, deterioration may be rapid shortly after presentation, particularly when the first sign is visual loss, and death generally occurs 4 to 6 years after onset; the median age at death was eight years after symptom onset in the group with onset between 13 and 36 months [2]. Above six years, initial manifestations may be behavioural (attention deficit hyperactivity disorder, mood disorders), followed by motor difficulties and rapid decline [2].

In the prospective cohort of Bascou et al. (35 children with onset between 6 and 36 months assessed between 2000 and 2017 at a single US centre), median age at onset was 11.5 months and median diagnostic delay was 3.5 months. Among the 32 symptomatic patients, the most frequent initial signs were loss of acquired skills (41%), irritability (38%), abnormal gait (22%), motor delay (16%) and abnormal tone (13%). During the course, 72% had feeding difficulties, 69% constipation, 53% reflux, 91% axial hypotonia, 88% appendicular hypertonia, 31% seizures and 47% frequent staring episodes. MRI was abnormal in 33 of 34 patients, nerve conduction in 21 of 24 (88%), brainstem auditory evoked potentials in 77% and visual evoked potentials in 42%; median CSF protein in symptomatic patients was 126 mg/dL. Median survival estimated by the Kaplan–Meier method was 6.72 years (95% confidence interval [CI]: 4.26 to infinity). The authors proposed a severity index with ten signs (reflux, constipation, feeding difficulties, impaired visual tracking, abnormal pupillary responses, staring, clonus, spasticity, seizures and abnormal reflexes) for the group with onset between 9 and 12 months, in which a score of 6 or more predicts the severe form [15]. These figures come from a referral centre and a predominantly white population; they do not represent population frequencies.

In adolescence and adulthood, the disease may present with loss of manual dexterity, tremor, burning paraesthesia and weakness without intellectual deterioration, with unilateral upper limb weakness and lower limb hypoaesthesia, or with slowly progressive spastic paraparesis; other patients become bedridden and deteriorate physically and mentally [2]. The adult group includes people with subtle symptoms since childhood diagnosed late — a woman described as “tremulous” since childhood, with a slow, stiff, broad-based gait, generalised deterioration after 40 years and death at 73 — and people entirely normal until after 20 years, such as a woman with spastic paraparesis from 38 years and demyelination confined to the corticospinal tract [2]. In the series by Debs et al. (11 patients and 30 literature cases, diagnosed after 16 years), 96% of the adult-onset group had pyramidal signs and all progressed to spastic paraparesis or tetraparesis; neuropathy occurred in 59% (demyelinating in 80% of these), dysarthria in 31%, cerebellar ataxia in 27%, pes cavus in 27%, deep sensory signs in 23%, tongue atrophy in 15%, optic neuropathy in 12% and cognitive decline in 12%; CSF protein was moderately elevated in 54% [23]. Saudubray summarises that most adults present with spastic paraparesis, with or without neuropathy, one third have cerebellar ataxia and cognitive dysfunction is generally absent [5]. In Brazil, the medical genetics group at Hospital de Clínicas de Porto Alegre described a 28-year-old woman with progressive bilateral amaurosis since 8 years, weakness of all four limbs and preserved intellect, homozygous for a novel missense variant with low but non-zero residual activity [24].

3.4. Manifestations by system

4. Clinical Recognition and Warning Signs

In a previously healthy infant, the combination of extreme irritability and unexplained crying, feeding difficulties with reflux, clenched fists beyond three months, lower limb rigidity with axial hypotonia and loss of acquired skills such as smiling, babbling and head control should prompt consideration of Krabbe disease [2]. Bone pain crises, expressed through irritability and incessant crying, may precede psychomotor deterioration with peripheral neuropathy by several weeks [25]. Saudubray includes neonatal Krabbe disease among causes of severe hypertonia with signs of hyperexcitability in newborns and lists infantile disease among causes of regression with painful pyramidal hypertonia and opisthotonus, optic atrophy with incessant crying and basal ganglia calcifications on computed tomography [25]. The Oxford Desk Reference recalls that the infantile form may manifest with dystonia (“lead-pipe” rigidity and abnormal postures), irritability, poor feeding, motor regression and seizures, and that MRI may show white matter abnormalities and calcifications, particularly of the basal ganglia [26].

Some signs are particularly valuable because they are uncommon in other leukodystrophies: the association of leukodystrophy with demyelinating peripheral neuropathy (shared with metachromatic leukodystrophy and some other diagnoses), elevated CSF protein without pleocytosis, absence of visceromegaly and thickening of the optic nerves or cauda equina roots [1, 6, 25, 27]. In older children and adults, warning signs are progressive spastic paraparesis with or without neuropathy, isolated visual loss and leukoencephalopathy following the corticospinal tracts and optic radiations [2, 5]. GeneReviews recommends assessing GALC activity in anyone, at any age, with progressive deterioration of the central or peripheral nervous system [2].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria

GeneReviews distinguishes two scenarios. In symptomatic patients, the biochemical diagnosis is established in the presence of a presentation compatible with the infantile form, elevated dried blood spot psychosine and deficient leucocyte GALC activity; activity is very low (0–5% of normal) in all symptomatic patients, and molecular confirmation is recommended in all. Molecular diagnosis requires biallelic pathogenic or likely pathogenic GALC variants; a heterozygous variant of uncertain significance (VUS) neither establishes nor excludes the diagnosis [2]. In asymptomatic newborns, diagnosis may be established by low GALC activity, elevated psychosine in the screening sample and biallelic pathogenic variants [2]. When VUS are identified, psychosine in the infantile range with low GALC suggests the infantile form [2].

Three limitations govern interpretation. GALC activity does not distinguish the infantile form from later-onset forms; low activity may result from pseudodeficiency alleles, heterozygosity for a pathogenic variant, variants associated with later onset or saposin A deficiency; and psychosine may decrease in advanced stages of the infantile form and is also elevated in saposin A deficiency [2]. The GeneGuidance curation, based on the 2018 GeneReviews version, adds that entirely normal leucocyte activity excludes the disease, but that elevated psychosine with normal GALC should prompt consideration of saposin A deficiency [28].

5.2. Biochemical tests

GALC activity. This can be measured in leucocytes, dried blood spots or cultured fibroblasts [5]. The assay is subject to technical (substrate) and biological (pseudodeficiency) errors. The radiolabelled natural substrate, the historical standard, is being replaced by liquid chromatography–tandem mass spectrometry (LC-MS/MS) techniques using short-chain analogues, and the less sensitive fluorogenic substrate should be used cautiously [5]. The Atlas emphasises that the natural-substrate assay is demanding and should be performed in an experienced laboratory [1]. In saposin A deficiency, GALC is decreased in leucocytes but not in cultured fibroblasts [5].

Psychosine. Escolar et al. (2017) established, in 75 normal newborns, a dried blood spot range of 0.14–0.53 nmol/L (mean 0.31) and an upper normal limit of 0.71 nmol/L. In neonatal samples from six children who developed the early-infantile form, psychosine ranged from 5.2 to 44 nmol/L; nine screen-positive newborns who did not develop disease had a median of 0.98 nmol/L (0.21–2.7). In this cohort, values above 3 nmol/L were 100% specific for infantile forms, and values between 0.7 and 3 nmol/L neither confirmed nor excluded disease. Variability between punches from the same sample was 12.1% (coefficient of variation), so the authors recommend analysing multiple punches; psychosine decreased with natural disease progression and after transplantation [29]. GeneReviews and the 2026 recommendations adopt an operational threshold of ≥10 nmol/L in dried blood spots for severe infantile disease, but GeneReviews records rare infantile cases with intermediate values (2–10 nmol/L), presented as a conference communication in 2026, and therefore recommends surveillance and anticipatory guidance during the first year [2, 6]. Values between 2 and 9 nmol/L identify children at risk of disease after 12 months, but also some who will not develop disease in childhood [6]. For confirmation, the 2026 recommendations prefer erythrocyte psychosine normalised to haemoglobin, because dried blood spot psychosine decreases with haematocrit after birth; the infantile form has ≥10 nmol/L in dried blood spots and ≥50 pmol/g haemoglobin in erythrocytes, and the units for the two matrices are not interchangeable [6].

Cerebrospinal fluid. Protein is elevated from stage I of the infantile form and inconsistently in later-onset forms [2, 5]. GeneReviews gives a normal range of 48–72 mg/dL in infants and 18–58 mg/dL in adults, recalling that ranges vary with age and laboratory [2]. Electrophoresis shows increased albumin and α-globulin and decreased β- and γ-globulins, a pattern also seen in metachromatic leukodystrophy [1].

5.3. Newborn screening

Screening measures GALC activity in dried blood spots collected between 24 and 72 hours of life, using tandem mass spectrometry (most programmes, often in a multiplex lysosomal disease panel) or fluorimetry (Missouri) [2, 10]. Regardless of method, activity alone is insufficiently specific because it overlaps between patients, carriers and people with pseudodeficiency [10]. Table 5.1 summarises the available evidence.

The 2018 guidance recommends second-tier testing — preferably psychosine, with or without complete molecular analysis — and referral of probable cases by days 5–7 of life (grade C and D recommendations because of the absence of trials) [10]. The infantile form was included in the RUSP in 2024, using the criterion “low GALC and psychosine ≥10 nM”, following rejections in 2010 and 2023; screening for later-onset forms was not recommended [8, 9]. The 2026 recommendations give the date as 1 July 2024 and, in November 2025, list 16 states besides New York with screening (Missouri, Kentucky, Ohio, Tennessee, Illinois, Indiana, New Jersey, Pennsylvania, Georgia, South Carolina, Minnesota, Maryland, Texas, Iowa, Oregon and New Mexico); most states using psychosine adopted a threshold of ≥2 nmol/L, which also detects later-onset forms [6]. The federal advisory committee was disbanded in March 2025, and the RUSP page remains as an archive [8].

The advisory committee (ACHDNC) documents, obtained from archived copies of HRSA files, clarify this pathway. In February 2023, following an evidence review for forms with onset up to 36 months, the vote was tied and inclusion was not recommended; the chair's letter to the nominators (22 March 2023) requested, for reconsideration, data systematically reporting transplant outcomes in early-infantile disease with appropriate case classification (psychosine or genetic markers), data on transplantation harms in the first two months of life, including in infants not meeting infantile criteria, and data on infants at risk of late-infantile disease, including losses to follow-up, the burden of intensive follow-up and consequences of indeterminate results [32]. That review projected, for all US newborns, approximately 70 infants per year with disease or at risk: approximately 15 infantile cases detected in time for treatment, with approximately 10 early-childhood deaths prevented, and approximately 22 high-risk and 33 low-risk infants for other forms requiring specialist follow-up, some of whom would never be diagnosed with disease [33]. The July 2023 renomination was restricted to the infantile form, with psychosine ≥10 nM as the referral criterion, and in August 2023 the committee approved an expedited review, the first for a reconsidered condition, covering literature from 10 January to 22 November 2023, data from the 9 state programmes with second-tier psychosine and conference abstracts not yet published [34].

In that review, the technical panel considered the risk of infantile disease with psychosine <10 nM to be almost zero: in 26 infantile cases psychosine ranged from 10 to 108 nM (median 48.5 nM), and in 11 later-onset cases only one had ≥10 nM (12 nM, at 460 days), with the remainder between 2.1 and 9.7 nM (median 3.9 nM) [34]. Uncertain exceptions were recorded: a conference abstract with infantile cases and neonatal psychosine of 2–6 nmol/L in samples collected since 2010 (possible degradation), an infant with 1.2 nM, initially classified as later-onset, who died at 26 months with an atypical course, and twins with 5.2 and 4.9 nM transplanted around 100 days and classified by the programme as late-infantile [34]. In the 9 programmes (3,551,033 screened), 1,505 had low GALC (42.4 per 100,000), 13 psychosine ≥10 nM and 11 infantile disease (3.1 per million); the only false positives were two samples contaminated with psychosine standard at a reference laboratory, with 3 days of hospitalisation, and 33 known or suspected later-onset cases (9.3 per million) would not have been referred using that threshold [34]. In the decision model for an annual cohort of 3.65 million, screening would identify 11.3 infantile cases (5.6–20.2), 9.9 would undergo transplantation, 1.0 would die from complications within the first 100 days and 2.1 would die by 30 months, compared with 7.9 deaths with clinical detection (24.2 Krabbe cases, of which 11.3 infantile, and only 1.1 transplanted); the model was restricted to mortality by 2.5 years because of a lack of evidence on quality of life, neurological status or longer survival [34]. In January 2024 the committee voted in favour of including the infantile form with two-tier screening (low GALC and psychosine ≥10 nM), and the Secretary of Health added it in July 2024; at the report date, 11 states screened, including New Jersey and Ohio without second-tier psychosine [35]. Regarding families, the review found only an online survey of 170 affected people, with 97% supporting screening but an incalculable response rate, and a low-quality disparities study in which the proportion of Black or Asian patients was twice as high among screen-detected cases and the interval from first hospitalisation to diagnosis was 64 days in non-Hispanic white children versus 101 to 136 days in other groups, without statistical significance [34]. The sources consulted identified no population Krabbe screening programme in Europe, Portugal or Brazil; this absence from the sources does not demonstrate that pilot programmes do not exist.

5.4. Algorithm after positive screening

Table 5.2 summarises the pathway proposed by GeneReviews and the US recommendations.

In New York, retrospective application of the 2021 classification to 47 newborns initially classified as moderate or high risk would have reduced follow-up testing by 88% [13]. The 2018 recommendations considered that the specialist care centre need not perform MRI, lumbar puncture or conduction studies before referral because transplant centres perform them rapidly (grade D) [10]; a 2025 report in two newborns advocates early electrodiagnostic testing because it demonstrated severe mixed sensorimotor polyneuropathy before confirmation was available and helped counsel families [36].

5.5. Molecular testing and test limitations

GeneReviews proposes GALC sequencing with targeted testing for the 30 kb deletion as the first step and, if only one or no variant is found, gene-targeted deletion/duplication analysis; multigene panels must include testing for the 30 kb deletion; and, if exome or genome sequencing is used, the deletion may escape exome detection, so genome sequencing or exome sequencing with targeted deletion analysis is recommended, preferably on a rapid basis [2]. The GeneReviews testing table gives the proportion of pathogenic alleles detected by each method: 55–65% by sequencing and 35–45% by testing for the 30 kb deletion; other single- or multiexon deletions are rare [2]. These proportions concern alleles and do not equal the clinical sensitivity of testing in an individual patient. Detection of deletions through exome or genome read depth may be less sensitive than targeted methods [2]. In the Kentucky case described by Kwon et al., standard sequencing showed a single variant and a novel deletion was found only through array comparative genomic hybridisation; elevated psychosine (61 nmol/L) enabled referral on day 6 and transplantation on day 24 [10].

Interpretation is difficult because genotype–phenotype correlations are lacking for most combinations and pseudodeficiency alleles are frequent; in January 2026, 402 of 1577 GALC variants in ClinVar (25%) were classified as VUS [6]. The transplantation decision should not be delayed by ambiguous genetic results when GALC and psychosine are compatible with infantile disease [6]. The report should include the gene, versioned transcript (NM_000153.4), HGVS nomenclature, zygosity, allele phase (parental testing), classification according to ACMG/AMP criteria and methodological limitations; it should also mention pseudodeficiency and modifying variants, which may not be reported in carrier testing because they are benign individually [2, 19]. Apparent homozygosity may represent an undetected deletion on one allele or uniparental isodisomy [2].

5.6. Imaging and electrophysiology

In the infantile form, MRI in the first months shows demyelination with T2 hyperintensity of periventricular white matter, deep grey matter and the centrum semiovale, cranial nerve enhancement and, on spinal MRI, nerve root enhancement; subcortical U fibres may be spared until late [2]. The Oxford Desk Reference describes, in the early-infantile form, hyperintensity in cerebellar white matter, the brainstem and internal capsule, with streaked periventricular and deep hyperintensity [27]. The 2026 recommendations state that, up to 4 months, the earliest abnormalities are located in cerebellar grey and white matter (including the dentate nuclei) and pyramidal tracts, and may involve parieto-occipital white matter and the splenium of the corpus callosum [6]. Imaging may be normal early in the disease [1], and the GeneGuidance curation, based on the 2018 GeneReviews version, adds that before 6 months MRI may appear misleadingly normal because of poor grey–white matter contrast [28]. Diffusion tensor imaging is the preferred modality for assessing progression; midbrain atrophy grading parallels overall progression, and brainstem atrophy is associated with apnoea and thermal instability [2]. Computed tomography shows diffuse cerebral atrophy with white matter hypodensity and may show thalamic, basal ganglia and periventricular calcifications [5]. In later-onset forms, lesions are more localised, with less cerebellar and deep grey matter involvement; in adults, hyperintensity along the pyramidal tracts, involving the optic radiations and corticospinal tracts, is almost constant [5]. In the series by Debs et al., corticospinal tract hyperintensity occurred in 94%, optic radiation hyperintensity in 89% and posterior corpus callosum abnormalities in 60% [23].

Motor nerve conduction is consistently reduced in the infantile form and most late-infantile forms, and only approximately 60% of juvenile and adult patients have signs of neuropathy [5]. In a series cited by the Atlas, 82% of 27 patients aged between 1 day and 8 years had uniform slowing of sensory and motor conduction [1]. Brainstem auditory evoked potentials show reduced amplitude, prolonged latencies and absence of waves III and V; visual evoked potentials are delayed or reduced; the electroencephalogram is initially normal and becomes slow and disorganised, sometimes asymmetrical [2]. Neural tissue biopsy is not indicated for diagnosis [4].

6. Differential Diagnosis

A history of normal development in the first months followed by psychomotor deterioration distinguishes the infantile form from non-progressive diseases of congenital or perinatal origin; later-onset disease is harder to distinguish from other degenerative diseases and may present solely with progressive visual loss [2]. The Oxford Desk Reference recommends that investigation of any leukodystrophy be guided by the MRI pattern assessed by an experienced paediatric neuroradiologist, distinguishing dysmyelinating disorders (such as Krabbe disease, metachromatic leukodystrophy, X-linked adrenoleukodystrophy, Alexander disease, Canavan disease, Aicardi–Goutières syndrome and vanishing white matter disease) from hypomyelinating disorders (such as Pelizaeus–Merzbacher disease), and recalls that a precise aetiology is not identified in approximately half of children with leukodystrophy, although panels, exome or genome sequencing enable a clinical diagnosis in 42% in one series [27]. Acquired leukoencephalopathies — hypoxic–ischaemic injury and congenital infection, particularly cytomegalovirus — and genetic disorders that mimic them (for example, COL4A1) also enter the differential [27]. In adults, a previous diagnosis of multiple sclerosis is a recognised error, as illustrated by the case in section 10 [20].

In adults, Saudubray includes Krabbe disease among sphingolipidoses presenting as spastic paraparesis with leukoencephalopathy specific to the corticospinal tracts, and among metabolic causes of demyelinating and axonal neuropathy and optic nerve injury; the adult approach tables place metachromatic leukodystrophy, adrenoleukodystrophy, cerebrotendinous xanthomatosis and peroxisomal diseases alongside each other [37]. Orphanet adds prosaposin deficiency encephalopathy to the differential [4].

7. Treatments and Risk Management Strategies

7.1. Principles

There is no cure or authorised medicine for Krabbe disease. Allogeneic haematopoietic stem cell transplantation is the only disease-modifying therapy available in clinical practice; it acts after engraftment and donor cell expansion, and must therefore be performed before rapid demyelination begins [2, 6]. Management depends on three variables: the predicted form (infantile or later-onset), clinical status at the time of the decision (asymptomatic, minimally symptomatic or symptomatic) and the family's informed choice, which may legitimately favour supportive care alone [6, 9]. Supportive and palliative care should be ensured in every scenario: for those ineligible, those declining transplantation and those who progress afterwards [6].

7.2. Transplantation in the infantile form

Evidence. In the study by Escolar et al. (2005), 11 asymptomatic newborns (12 to 44 days), diagnosed through family history, and 14 symptomatic infants (142 to 352 days) received unrelated donor umbilical cord blood after myeloablative chemotherapy. Engraftment and survival were 100% and 100% in asymptomatic patients (median follow-up 3.0 years) and 100% and 43% in symptomatic patients (3.4 years). Asymptomatic patients had progressive central myelination and developmental gains, with age-appropriate cognition and receptive language in most, but some mild to moderate expressive language delays and mild to severe gross motor delays; symptomatic patients had minimal neurological improvement [17]. Follow-up of these patients showed that most subsequently developed speech and motor difficulties [2]. A staging system based solely on signs and symptoms, developed in 42 patients assessed for transplantation, predicted post-transplant development in 29 transplant recipients; neurophysiological and imaging tests added no value to the algorithm [38].

The 15-year study by Wright et al. (2017) followed 18 children with infantile disease transplanted in the first seven weeks (mean follow-up 9.5 years, 4–15): five died (three from peritransplant complications, one from an unrelated surgical complication and one from progression), one survivor had normal cognition and ten continued acquiring cognitive skills at a slightly slower rate; all gained receptive language, seven within the normal range; three walked without support and seven with devices; 12 used orthoses; fine motor function was generally preserved; myelination and atrophy stabilised in eight, improved in four and worsened in one; nerve conduction initially improved but remained abnormal in most. The study provides class IV evidence [39]. GeneReviews adds that motor outcome reflected the severity of corticospinal involvement already present at birth, that peripheral nerve disease progressed with severe muscle atrophy and scoliosis and that, into adolescence, treated children may have relatively normal lives with variable motor disability [2]. The ten patients transplanted in the first four weeks had better survival and function than those transplanted in the second month [10], and Allewelt et al., according to the 2026 recommendations, described more favourable mobility, communication and feeding outcomes when transplantation occurred before 30 days [6].

These data do not allow direct comparison between centres or periods because age at transplantation, clinical status, donor type and conditioning regimen differ; the meta-analysis showed an inverse association between reported survival and study size/quality, suggesting overestimation in small series, and the authors caution that survival does not equal preservation of neurological function [40]. The 2026 recommendations summarise accumulated experience: transplant-related mortality of 10–15%, overall 10-year survival of 75–85%, independent walking in 20–50% of presymptomatic transplant recipients, gastrostomy dependence in almost 60% of long-term survivors, relatively preserved receptive language and non-verbal cognition and delayed or limited expressive language [6]. The review for the RUSP estimates a death risk of approximately 10% within the first 100 days [9]. Already symptomatic infants are not candidates [6].

Timing window. The 2018 guidance recommends referral to an experienced centre by the third week of life to initiate transplantation in the first four weeks (grade D) [10]. The review for the RUSP states that transplantation must be completed before 4–6 weeks, and official HRSA summaries recommend it in the first 6 weeks of life, before major signs [9, 35]; the 2026 recommendations specify ideally by 30 days [6]; and GeneReviews, revised in September 2026, refers to treatment “ideally no later than day 14 of life” [2]. The targets do not contradict the principle — the earlier, the better — but differ numerically; in practice, the operational goal of US programmes is transplantation in the first month, and the series by Hall et al. shows transplantation on average by day 36 with two-tier screening [31].

7.3. Transplantation in later-onset forms

In later-onset forms, slower progression allows benefit if diagnosis is sufficiently early [2]. In 19 patients with onset between 13 and 36 months (Yoon et al., 1997–2020), transplantation prolonged survival and improved cognition and language compared with untreated patients; asymptomatic patients had normal or near-normal development in all domains, and among symptomatic patients, those with onset after 12 months had better cognitive outcomes than untreated patients, whereas those with onset at or before 12 months resembled untreated patients [22]. In a report cited by GeneReviews, transplantation in a later-onset form halted demyelination and axonal loss [2]. In six transplanted Italian juvenile patients (four symptomatic, two presymptomatic; follow-up from 9 years and 2 months to 19 years and 8 months), all survived, four had near-normal cognitive, motor and functional status, and the two with extensive white matter disease, epilepsy and cognitive impairment before transplantation had suboptimal outcomes; GALC activity normalised in all [41]. The Atlas and Saudubray record stabilisation of some later-onset patients following bone marrow or cord blood transplantation [1, 5]. The review for the RUSP considers information on the impact of transplantation in later-onset forms insufficient, regardless of timing [9], and the 2018 consensus recalls that longitudinal studies are needed to determine whether psychosine can guide transplantation timing in these children [10].

7.4. Transplantation organisation

The guidance describes a process that must fit within a few weeks: admission to the transplant centre around days 7–8, urgent high-resolution HLA typing, psychosine, CSF, MRI with diffusion tensor imaging, electroencephalogram, evoked potentials and nerve conduction, hearing and visual assessment, echocardiography (for patent ductus arteriosus or foramen ovale, with filters on intravenous lines), cord blood unit selection, central catheter placement and possible gastrostomy, conditioning chemotherapy (nine days at the time) and transplantation around day 21 or later [10]. Cord blood is the most commonly used source because of rapid availability, the potential for central nervous system engraftment and the lower risk of chronic graft-versus-host disease; a sibling donor must be a non-carrier, with normal activity and HLA compatibility, and parental haploidentical donation is inappropriate because parents are carriers [6]. The 2026 recommendations propose starting chemotherapy while confirmation of the cord unit and a backup unit is underway, differing from traditional protocols requiring prior confirmation [6].

Complications include graft-versus-host disease (more frequent with HLA mismatch), seizures, encephalopathy, posterior reversible encephalopathy syndrome, central nervous system infections and organ toxicity from myeloablative regimens; all newborns undergoing chemotherapy will have dental abnormalities, and myeloablative and reduced-intensity regimens have not been sufficiently compared [6]. After transplantation, general surveillance protocols for paediatric transplant complications apply alongside specific neurological follow-up, whose precise protocol is not established [6]. Psychosine decreases after transplantation but remains above normal in most infantile forms [29].

7.5. Supportive treatment and situations to avoid

The sources consulted do not give doses for symptomatic medicines; these should follow paediatric protocols and each medicine's label, and none has a specific indication for Krabbe disease. The medicines listed below are used for symptoms (spasticity, seizures, pain, reflux), not for the disease itself.

GeneReviews advises avoiding atypical antipsychotics and antiseizure polytherapy, which may cause excessive sedation, depress respiratory drive and, in the authors' experience, accelerate neurodegeneration; avoiding routine vaccines, including live viral vaccines, because the immune response may accelerate progression (except annual influenza vaccination); avoiding long-term indwelling catheters; and avoiding extreme temperatures, dehydration, untreated infections and severe reflux, which may trigger seizures [2]. These recommendations are based on the authors' personal observations and a clinical review, rather than trials, and apply to symptomatic patients [2, 42].

7.6. Regulatory status

This review did not verify the status of FDA orphan designations or national access in Portugal (INFARMED) or Brazil (ANVISA/CONITEC); as no medicine is authorised, access mainly concerns transplantation and screening.

7.7. Clinical trials and research

The 2026 recommendations report that two gene therapy trials enrolled several infants but were closed by the companies following changes in commercial priorities, and that new trials are being planned; they also mention molecules under study that reduce psychosine (UGT8 inhibitor) or promote myelination (clemastine and Sob-AM2) [6]. Commercial failures do not equal negative trials: no published efficacy results permit that conclusion. Preclinical research published in the last 12 months in the twitcher mouse includes intracranial AAV9-GALC monotherapy targeting the thalamus and deep cerebellar nuclei, with normalisation of psychosine and near-normal life expectancy [50]; adenine base editing using two intracerebroventricular AAV9 vectors, which corrected approximately 0.5% of genomic DNA, restored approximately 5% of activity and prolonged survival [51]; and ex vivo lentiviral stem cell gene therapy using a chimeric GALC enzyme designed for improved secretion and blood–brain barrier crossing [52]. GeneReviews also mentions animal studies of enzyme replacement therapy, neural stem cell transplantation, substrate reduction and chemical chaperones, and synergy between transplantation and gene therapy in murine and canine models [2]. None of these approaches has, to date, demonstrated efficacy in humans.

8. Reproductive Options and Family Counselling

Inheritance is autosomal recessive. A patient's parents are presumed heterozygous, and parental molecular testing is recommended to confirm carrier status and variant phase. If only one parent has the variant and paternity and maternity are confirmed, a de novo variant in the patient or parental postzygotic mosaicism should be considered; if the patient appears homozygous, a deletion undetected by sequencing or uniparental isodisomy should be considered [2]. Heterozygotes are asymptomatic and not at risk of developing Krabbe disease [2].

When both parents are carriers, each sibling of a patient has, at conception, a 25% probability of being affected, 50% of being an asymptomatic carrier and 25% of being neither affected nor a carrier [2]. Children of a person with a later-onset form are obligate carriers, and each sibling of the patient's parents has a 50% probability of being a carrier [2]. Intrafamilial variability in later-onset forms prevents using one sibling's age at onset to predict another's, which is why the genetic status of older siblings, even asymptomatic ones, should be clarified to enable timely transplantation [2]. The case in section 10.1, in which an adult's diagnosis led to diagnosis of two siblings, illustrates this rule [20].

Carrier detection. Reliable carrier testing requires prior identification of the familial variants; GALC activity measurement cannot determine carrier status because carrier values overlap with normal values and those of pseudodeficiency alleles [1, 2]. The Oxford Desk Reference summarises that familial enzyme testing rarely discriminates sufficiently to test carriers' spouses [27]. Carrier testing should be considered for partners of known carriers and people with later-onset disease, particularly if they share ancestry; a founder variant exists in the Druze population [2, 53]. Expanded carrier tests may not report pseudodeficiency or modifying variants because they consider them benign, which, according to Schecter et al., is relevant to family planning when a haplotype such as p.Tyr319Cys with p.Ile562Thr in cis is present [19].

Prenatal and preimplantation diagnosis. Once familial variants are identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible [2, 4]. Prenatal diagnosis was historically performed by enzyme assay in villi and amniocytes, with the recommendation to measure parental activity beforehand to avoid false positives due to pseudodeficiency or heterozygotes with very low values; it is now mainly performed through molecular genetics [1, 5]. Prenatal diagnosis also allows preparation for transplantation immediately after birth [2]. The decision to use these options is personal and should be discussed before pregnancy without directive pressure [2].

Counselling after positive screening. Communication with the family should explain enzyme deficiency and psychosine, the untreated natural history, what transplantation can and cannot achieve, the urgency and the legitimacy of choosing supportive care; the 2026 recommendations propose model wording for each objective, including that “10–15% of babies will not survive transplantation, but 85–90% will survive” and that transplantation is a treatment rather than a cure [6]. In US experience, approximately 30% of families of newborns with infantile disease declined transplantation [9]. Children classified as at risk of later-onset disease create a different problem: prolonged uncertainty, family anxiety and poor adherence to follow-up; in New York, none of approximately 40 at-risk children, with a median follow-up of 5 years, had concerning symptoms at the time of the 2018 guidance [10]. DNA banking is advised when the molecular diagnosis has not been confirmed [2].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Global, American and European frequency

Estimates depend on the method. GeneReviews reports approximately 1 in 100,000 births, with median prevalence varying greatly between countries, and Orphanet gives an estimated prevalence of 1/100,000 in northern Europe and worldwide incidence of 1/100,000 to 1/250,000 live births, with the infantile form accounting for 85–90% of cases in northern Europe and later-onset forms possibly more frequent in other populations [2, 4]. Saudubray gives a global incidence between 0.75 and 1 per 100,000 live births, higher in Scandinavia (but not Finland), and considers the incidence of later-onset forms underestimated, these being more common in Japan, Italy and Sicily [5]. The review for the RUSP, based on clinical detection in other countries, presents a range of 0.3 to 2.6 per 100,000 live births and an estimate of 30% to 70% for the proportion with onset in the first six months [9].

These frequencies measure different things. Hospital series, such as the Sicilian or Danish series, describe the composition of diagnosed patients rather than population prevalence; the US mortality rate measures coded deaths; and New York screening incidence applies to confirmed infantile disease, not to all newborns with low GALC and two variants, whose course is uncertain [7, 54]. GALC-related disease frequencies do not include saposin A deficiency, a distinct entity with four known cases according to Saudubray [5]. The Atlas also reports that, in an Israeli series, all patients were Arab, and parental consanguinity was observed; the Oxford Desk Reference lists the disease as very rare, more frequent in Druze and Sicilian populations [1, 53].

9.2. Portugal and Brazil

The sources consulted yielded no epidemiological data from Portugal or Krabbe newborn screening programme in Portugal or Brazil; this absence from the sources consulted does not demonstrate the absence of studies. In Brazil, the main literature includes the adult case described by the Medical Genetics Service at Hospital de Clínicas de Porto Alegre, with a prolonged course and a novel homozygous variant [24], and Porto Alegre was among the centres of the currently suspended PBKR03 gene therapy trial [46].

9.3. Special populations

Transplantation in newborns raises specific challenges: the infant's health insurance must be established rapidly, transfer to another state or region may be necessary, and experienced centres are few, creating geographical and financial barriers [6, 10]. In adolescent and adult forms, diagnosis takes approximately 10 years on average in the Debs and Kolodny series, according to Almudhry et al., and may be preceded by alternative diagnoses such as multiple sclerosis [20]. The sources consulted yielded no specific data on pregnancy in women with the adult form.

9.4. Referral networks and patient associations

In Europe, leukodystrophies are managed through the European Reference Network for Rare Neurological Diseases (ERN-RND), whose leukodystrophies page provides centres and contacts, standards of care in several languages, patient information, patient organisations, trials and registries [14]. In the USA, the Leukodystrophy Care Network, sponsored by the Hunter's Hope association, certifies centres of excellence and hosts the monthly Krabbe Disease NBS Council, which discusses new cases; the 2026 recommendations also mention KrabbeConnect [6]. GeneReviews lists Partners for Krabbe Research, The Rosenau Family Research Foundation, United Leukodystrophy Foundation, the Myelin Disorders Bioregistry Project, HRSA's state newborn screening page and the Pittsburgh Program for Neurodevelopment in Rare Disorders registry [2]. A global Krabbe patient registry (NCT02993796) exists and is recruiting [55].

10. Illustrative Clinical Case

10.1. Later-onset disease in siblings, with a previous diagnosis of multiple sclerosis

Almudhry et al. described, in 2025, two siblings of Middle Eastern origin, with no known parental consanguinity, followed in a neurometabolic disease clinic in London, Ontario (Canada) [20]. The proband had unexplained dysphagia episodes at 8 years and, at 15, symmetrical ascending lower limb numbness up to T4. The clinical presentation and MRI led to a diagnosis of paediatric-onset multiple sclerosis at another tertiary centre. Sensory symptoms relapsed, diplopia and motor weakness developed, the response to corticosteroids was inconsistent and several disease-modifying multiple sclerosis therapies provided no benefit. Over the following decade he began using a walking stick and developed incontinence, erectile dysfunction, word-finding difficulty, poor concentration and articulation abnormalities, but completed a university postgraduate qualification [20].

At 32 years, at the first neurometabolic assessment, he had dysarthria, asymmetrical spastic paraparesis, diffusely reduced tendon reflexes with bilateral Babinski signs, posterior column sensory impairment with hyperaesthesia, unstable gait with bilateral foot drop and limb ataxia. CSF protein was 1600 mg/L (reference 200–400) without pleocytosis, with a normal IgG index and negative oligoclonal bands; nerve conduction showed moderately severe demyelinating sensorimotor neuropathy. MRI showed confluent hyperintensity along the corticospinal tract, from the motor cortex to the internal capsule, and asymmetrical periventricular foci in the white matter and corpus callosum (Figure 2 of the article). GALC sequencing identified NM_000153.4:c.857G>A p.(Gly286Asp), classified as pathogenic in nine ClinVar submissions and likely pathogenic in one, and c.1034-2A>G, which alters the intron 9 acceptor, likely pathogenic; no measurable leucocyte β-galactocerebrosidase activity was present [3, 20]. The pedigree, with three affected siblings, appears in Figure 1 of the article [20].

The proband's diagnosis prompted family testing. His older sister also had later-onset disease, deteriorated rapidly and died before she could undergo transplantation. The younger brother was assessed at 21 years, asymptomatic, with normal MRI, and initially declined molecular testing; at 24 years he developed nocturnal numbness of some fingers and cramps, with minimal asymmetrical pyramidal signs, reduced vibration sense, mild demyelinating neuropathy, subtle corticospinal tract hyperintensities and GALC activity of 0.6 nmol/17 h/mg protein (reference 18–77), with the same genotype [20].

The younger brother underwent transplantation at 26 years (HLA 10/10 related donor) and the proband at 40 years (HLA 10/10 unrelated donor), both after conditioning with busulfan, cyclophosphamide and antithymocyte globulin, and graft-versus-host disease prophylaxis with methotrexate and tacrolimus; neither developed graft-versus-host disease. Enzyme activity increased to 28 and 12 nmol/17 h/mg protein, respectively. In the younger brother, neuropathy improved, examination normalised and MRI on day 677 remained stable, with persistent fatigue, difficulty concentrating, depressed mood and sphincter urgency. In the proband, four years later, disease had stabilised clinically and radiologically, but sensory, pyramidal, cerebellar and neuropathic deficits persisted: he stands with assistance and transfers to his wheelchair using a transfer board [20]. The conditioning regimens described concern these patients and do not constitute a general recommendation.

The case teaches four points: later-onset disease may mimic multiple sclerosis for years; the combination of central and peripheral involvement, progression without an immunotherapy response and a symmetrical corticospinal and optic radiation pattern should prompt GALC measurement; intrafamilial variability is substantial (the sister's death before treatment); and early transplantation while symptoms were still mild was associated in this brother with a better outcome than transplantation in advanced disease, although comparison between two patients does not permit causal inference [20].

10.2. Late-infantile form detected through newborn screening, with a pseudodeficiency and modifying haplotype

Schecter et al. described, in 2026, a boy screened in New York, born to first cousins of Southeast Asian ancestry, with dried blood spot GALC at 2.2% of the daily mean (controls >12%) and homozygosity for an in cis haplotype NM_000153.4:c.[742G>A;956A>G;1685T>C] p.[(Asp248Asn;Tyr319Cys;Ile562Thr)] (published on NM_000153.3, with identical numbering), alongside a benign promoter variant; both parents were heterozygous [19]. Leucocyte GALC was 0.03 nmol/h/mg protein (high-risk range 0.0–0.15). Assessment at five weeks (video electroencephalogram, MRI with spectroscopy and lumbar puncture) was normal except for slightly elevated CSF protein (67 mg/dL), and MRI at 15 and 27 months was also normal. The next MRI, delayed by the COVID-19 pandemic until 3 years and 9 months, showed parietal and periventricular FLAIR hyperintensity, involvement of the splenium, cerebral peduncles, midbrain, pons and cerebellar white matter, and cervical spinal cord abnormalities, with a score of 9 on the globoid cell leukodystrophy MRI scale; examination showed only mild appendicular hypotonia and erythrocyte psychosine was 14 pmol/g haemoglobin (controls <5) (Figure 2 of the article) [19].

The patient received cord blood transplantation shortly before 4 years, with normalisation of GALC activity and a fall in psychosine to 7 pmol/g; approximately 17 months later donor chimerism was lost, with autoimmune cytopenias and grade 2 gastrointestinal graft-versus-host disease, and a second cord transplantation was performed 18 months after the first, with full chimerism maintained for more than two years. At 9 years, cognitive, language and motor development was age-appropriate, without seizures and with stable MRI (Figure 4 of the article) [19]. Retrospective analysis of the neonatal sample showed slightly elevated psychosine (3.8, with controls <2.0), during a period when psychosine was not part of screening [19].

This case shows that a genotype composed of common variants of disputed significance may be associated with disease when combined in homozygosity, that intermediate neonatal psychosine does not exclude childhood disease and that imaging and psychosine surveillance enabled transplantation before any regression. The authors' proposal to reclassify p.Tyr319Cys as likely pathogenic and systematically test for p.Ile562Thr in cis is an interpretation based on one case and does not replace formal variant classification [19].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

Without treatment, the infantile form is uniformly fatal: mean age at death is 24 months (8 months to 9 years), most children become paraplegic in the second year of life, and the review for the RUSP reports median survival of 1.5 years in those with onset in the first six months [2, 6, 9]. The official 2024 HRSA summary describes four forms: infantile (onset at 0 to 12 months, usually around 4 months, mean survival approximately 1.5 years), late-infantile (13 to 36 months, around 14 months, approximately 9.5 years), juvenile (37 to 180 months, around 4 years, approximately 80% alive after 16 years) and adult (after 15 years, around 32 years, almost 90% alive after 19 years) [35]. In forms with onset between 12 and 36 months, the review for the RUSP reports median untreated survival of 9.5 years, GeneReviews a median age at death eight years after onset, and the cohort of Bascou et al. (onset between 6 and 36 months) median survival of 6.72 years [2, 9, 15]. Orphanet states that late-infantile/juvenile forms are generally fatal 2 to 7 years after onset and that adults may survive many years, some into the sixth decade with mild symptoms [4]. In the adult series by Debs et al., disease progressed slowly over more than 10 years, with rapid progression in two patients [23].

With presymptomatic transplantation, 10-year survival is between 75% and 85%, most retain relatively preserved cognition and receptive language, and motor disability and progressive peripheral neuropathy remain the main problem, potentially worsening in adolescence [2, 6]. Orphanet summarises that most children with predicted infantile disease treated before 4 weeks may live into the second decade with largely preserved cognitive function but difficulties with walking and expressive language [4]. Outcomes vary widely — from normal life with independent walking to tetraplegia — and depend on the earliness of detection, severity and progression before treatment [2]. A Japanese report from 2026 illustrates what “long survival” after transplantation means: an adolescent with infantile disease diagnosed at birth (because of an affected sibling), transplanted with unrelated donor cord blood at 41 days, underwent surgery for acetabular dysplasia at 5 years and progressive scoliosis at 13; at 16 years he measured 140 cm and weighed 21 kg, had contractures and marked lower limb weakness, used a wheelchair and was entirely dependent for daily activities, and underwent pulmonary wedge resection for recurrent spontaneous pneumothorax without complications; the authors consider the relationship between disease and pneumothorax uncertain [56]. Psychosine decreases with natural progression and transplantation, so low psychosine in a child with advanced symptomatic disease does not exclude disease [2, 29].

11.2. Surveillance of asymptomatic children at risk of later-onset disease

GeneReviews adapts the 2021 classification, separating asymptomatic newborns into high or low risk for later-onset disease; follow-up should be provided by a physician experienced in leukodystrophies, and the family should be educated to recognise signs [2, 13].

Table source: GeneReviews, tables 6a and 6b, adapted from Thompson-Stone et al. 2021 [2, 13].

11.3. Surveillance of symptomatic or transplanted patients

In symptomatic patients, GeneReviews recommends measuring growth and assessing nutritional status and oral feeding safety, signs of hydrocephalus, gait, mobility, seizures and movement disorders, development and educational needs, anxiety, attention deficit and mood, scoliosis and hip subluxation, vision and corneal ulcers, constipation, aspiration and respiratory failure at each visit; swallowing studies when indicated; annual densitometry (DXA) from 1 year in those bearing weight, until full weight-bearing, and in non-ambulant patients according to endocrinology; ophthalmological assessment according to the ophthalmologist; and the need for social support, palliative and respite care [2]. Initial assessment includes gastroenterology and nutrition, neurology, MRI with diffusion tensor imaging, nerve conduction, audiology with auditory evoked potentials, ophthalmology with visual evoked potentials, developmental and neuropsychiatric assessment (above 12 months), genetic counselling and family support [2]. After transplantation, general surveillance protocols for paediatric transplant complications and specific neurological follow-up of neuropathy, spasticity, pain, feeding and growth apply, although the exact protocol is not established [6]. Special education programmes and therapies should aim to maintain skills and prevent contractures and scoliosis, rather than improvement alone [2].

During transition to adulthood, GeneReviews recommends including a transition plan in the individualised education programme, and surveillance of at-risk children extends into adulthood at intervals of 2–5 years [2]. Quality-of-life and caregiver burden data after transplantation remain scarce; the 2026 meta-analysis emphasises the lack of standardised neurological, MRI and quality-of-life outcomes and proposes that future prospective studies and registries prioritise these [40].

12. Summary and Final Considerations

Krabbe disease results from biallelic GALC variants that reduce galactocerebrosidase activity; psychosine accumulation destroys oligodendrocytes and Schwann cells and produces central and peripheral demyelination, with globoid cells in the white matter [1, 5]. Saposin A deficiency, due to PSAP variants, is a rare phenocopy with low leucocyte GALC and elevated psychosine, which should be excluded when GALC genetics does not explain the presentation [2, 5].

The infantile form, the most frequent in clinical series from northern Europe and the USA, begins in the first months with irritability, feeding difficulties, hypertonia and loss of acquired skills and leads to death around two years; later-onset forms span the second year of life to the seventh decade, with spastic gait, visual loss, neuropathy and variable progression, and are more frequent in populations such as Japanese, Sicilian or Chinese populations [2, 5, 20]. Global frequency is approximately 1 in 100,000 births, with Druze and Arab founder effects in Israel and a Sicilian founder effect in Catania; the infantile form was detected in approximately 1 in 394,000 newborns screened in New York [2, 7].

Diagnosis rests on three pillars — GALC activity, psychosine and GALC genotyping including testing for the 30 kb deletion — and none is sufficient alone: activity neither distinguishes forms nor excludes pseudodeficiency, psychosine has intermediate zones and decreases with progression and transplantation, and a quarter of ClinVar variants are of uncertain significance [2, 6, 29]. Two-tier newborn screening, with psychosine ≥10 nmol/L as the RUSP criterion since 2024, identifies the infantile form with high specificity; lower thresholds detect later-onset forms for which the benefit of neonatal detection is not established [9, 31].

Haematopoietic stem cell transplantation, mainly using umbilical cord blood, performed before symptoms and ideally in the first month, increases survival and preserves much of cognition and receptive language, but does not prevent motor deficits, progressive neuropathy and gastrostomy dependence in many survivors, and carries associated mortality of 10–15% in reference series [6, 39]. In later-onset forms, benefit is greater when transplantation precedes symptoms or occurs with mild disease [22, 41]. No medicine is authorised; AAV vector gene therapy programmes were discontinued or suspended following corporate decisions, without published efficacy results, and preclinical research explores AAV9 vectors, base editing and ex vivo gene therapy with chimeric enzymes [6, 44, 46].

For the geneticist, the decision points are: treat screening with elevated psychosine as an emergency; do not delay transplant referral while awaiting ambiguous genetic results; document variant phase and pseudodeficiency alleles; clarify siblings' genetic status; monitor children at risk of later-onset disease at defined intervals; and support families, including those choosing supportive care, with palliative support and referral networks (ERN-RND in Europe) [2, 6, 14].

13. Bibliographic References

1. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Krabbe disease/galactosylceramide lipidosis/globoid cell leukodystrophy. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 93, p. 706–712. Local chapter.
2. Orsini JJ, Escolar ML, Wasserstein MP, Caggana M. Krabbe Disease. In: GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. First published 19 Jun 2000; full update 8 Sep 2026. NBK1238.
3. 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. Consultation of 15 variants of GALC (NM_000153.4, GRCh38, MANE Select) on 6 Oct 2026 via rest.variantvalidator.org.
4. Orphanet. Krabbe disease. ORPHA:487 [Internet]. Last updated Mar 2021. Disease page. Accessed 6 Oct 2026.
5. Vanier MT, Caillaud C, Levade T. Disorders of Sphingolipid Synthesis, Sphingolipidoses, Niemann-Pick Disease Type C and Neuronal Ceroid Lipofuscinoses. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 40, sections 40.2, 40.2.5 and 40.2.9. doi:10.1007/978-3-662-63123-2_40.
6. Thompson Stone R, White AL, Rubin JP, Rajan DS, Matern D, Kurtzberg J, et al. Expert Recommendations for Rapid Response to Positive Newborn Screen for Infantile Krabbe Disease. Neurol Genet. 2026;12(3):e200383. doi:10.1212/NXG.0000000000200383. PMID:42040243.
7. Orsini JJ, Kay DM, Saavedra-Matiz CA, Wenger DA, Duffner PK, Erbe RW, et al. Newborn screening for Krabbe disease in New York State: the first eight years' experience. Genet Med. 2016;18(3):239–248. doi:10.1038/gim.2015.211. PMID:26795590.
8. Health Resources and Services Administration. Recommended Uniform Screening Panel; Previously Nominated Conditions [Internet]. Rockville: HRSA; page revised Jul 2024 (archive after the committee was disbanded in Mar 2025). RUSP; Nominated conditions. Accessed 6 Oct 2026.
9. 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.
10. Kwon JM, Matern D, Kurtzberg J, Wrabetz L, Gelb MH, Wenger DA, et al. Consensus guidelines for newborn screening, diagnosis and treatment of infantile Krabbe disease. Orphanet J Rare Dis. 2018;13(1):30. doi:10.1186/s13023-018-0766-x. PMID:29391017.
11. U.S. Food and Drug Administration. openFDA Drug Label API: indication search using the term “Krabbe” [Internet]. api.fda.gov. Accessed 6 Oct 2026 (no results).
12. European Medicines Agency. Orphan designations (EU/3/21/2511, EU/3/21/2417, EU/3/11/911) [Internet]. EMA orphan designations. Accessed 6 Oct 2026.
13. Thompson-Stone R, Ream MA, Gelb M, Matern D, Orsini JJ, Levy PA, et al. Consensus recommendations for the classification and long-term follow up of infants who screen positive for Krabbe Disease. Mol Genet Metab. 2021;134(1–2):53–59. doi:10.1016/j.ymgme.2021.03.016. PMID:33832819.
14. European Reference Network for Rare Neurological Diseases (ERN-RND). Leukodystrophies — Disease Knowledge Hub [Internet]. ERN-RND page. Accessed 6 Oct 2026.
15. Bascou N, DeRenzo A, Poe MD, Escolar ML. A prospective natural history study of Krabbe disease in a patient cohort with onset between 6 months and 3 years of life. Orphanet J Rare Dis. 2018;13(1):126. doi:10.1186/s13023-018-0872-9. PMID:30089515.
16. Suzuki K, Suzuki Y. Globoid cell leucodystrophy (Krabbe's disease): deficiency of galactocerebroside beta-galactosidase. Proc Natl Acad Sci U S A. 1970;66(2):302–309. doi:10.1073/pnas.66.2.302. PMID:5271165.
17. Escolar ML, Poe MD, Provenzale JM, Richards KC, Allison J, Wood S, et al. Transplantation of umbilical-cord blood in babies with infantile Krabbe's disease. N Engl J Med. 2005;352(20):2069–2081. doi:10.1056/NEJMoa042604. PMID:15901860.
18. Fiumara A, Barone R, Arena A, Filocamo M, Lissens W, Pavone L, et al. Krabbe leukodystrophy in a selected population with high rate of late onset forms: longer survival linked to c.121G>A (p.Gly41Ser) mutation. Clin Genet. 2011;80(5):452–458. doi:10.1111/j.1399-0004.2010.01572.x. PMID:21070211.
19. Schecter DR, Donnelly C, White A, Raynes H, Heller G, Rajan D, et al. Homozygosity for a Clinically Significant GALC Haplotype Associated with Late-Infantile Krabbe Disease Detected on Newborn Screening: Implications for Clinical Management and Genetic Counseling. Int J Neonatal Screen. 2026;12(3):59. doi:10.3390/ijns12030059. PMID:42647246.
20. Almudhry M, Prasad C, Tay KY, Rupar CA, Atkins H, Prasad AN. Outcome of two siblings with late-onset Krabbe disease following allogeneic hematopoietic stem cell transplantation: And review of literature. Mol Genet Metab Rep. 2025;44:101242. doi:10.1016/j.ymgmr.2025.101242. PMID:40727947.
21. Spiegel R, Bach G, Sury V, Mengistu G, Meidan B, Shalev S, et al. A mutation in the saposin A coding region of the prosaposin gene in an infant presenting as Krabbe disease: first report of saposin A deficiency in humans. Mol Genet Metab. 2005;84(2):160–166. doi:10.1016/j.ymgme.2004.10.004. PMID:15773042.
22. Yoon IC, Bascou NA, Poe MD, Szabolcs P, Escolar ML. Long-term neurodevelopmental outcomes of hematopoietic stem cell transplantation for late-infantile Krabbe disease. Blood. 2021;137(13):1719–1730. doi:10.1182/blood.2020005477. PMID:33150395.
23. Debs R, Froissart R, Aubourg P, Papeix C, Douillard C, Degos B, et al. Krabbe disease in adults: phenotypic and genotypic update from a series of 11 cases and a review. J Inherit Metab Dis. 2013;36(5):859–868. doi:10.1007/s10545-012-9560-4. PMID:23197103.
24. Jardim LB, Giugliani R, Pires RF, Haussen S, Burin MG, Rafi MA, et al. Protracted course of Krabbe disease in an adult patient bearing a novel mutation. Arch Neurol. 1999;56(8):1014–1017. doi:10.1001/archneur.56.8.1014. PMID:10448809.
25. 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 1.3.4, 1.4.10 and 1.5 (tables 1.15 and 1.24). doi:10.1007/978-3-662-63123-2_1.
26. Firth HV, Hurst JA. Dystonia. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 139–140.
27. Firth HV, Hurst JA. Leukodystrophy/leukoencephalopathy. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 208–211.
28. GeneGuidance. Clinical curation of the GALC and PSAP genes (database export dated 26 Sep 2026; GALC entry based on GeneReviews NBK1238, 2018 version). Project database, accessed 6 Oct 2026.
29. Escolar ML, Kiely BT, Shawgo E, Hong X, Gelb MH, Orsini JJ, et al. Psychosine, a marker of Krabbe phenotype and treatment effect. Mol Genet Metab. 2017;121(3):271–278. doi:10.1016/j.ymgme.2017.05.015. PMID:28579020.
30. Wasserstein MP, Andriola M, Arnold G, Aron A, Duffner P, Erbe RW, et al. Clinical outcomes of children with abnormal newborn screening results for Krabbe disease in New York State. Genet Med. 2016;18(12):1235–1243. doi:10.1038/gim.2016.35. PMID:27171547.
31. Hall PL, White AL, Peck D, Pino G, Studinski A, Gavrilov D, et al. Two-tiered newborn screening for infantile Krabbe disease allows timely treatment initiation and avoids false-positive results. Genet Med. 2026;28(4):101572. doi:10.1016/j.gim.2025.101572. PMID:40888240.
32. Calonge N (Advisory Committee on Heritable Disorders in Newborns and Children). Letter to Hunter's Hope Foundation regarding the nomination of Krabbe disease for the RUSP. Rockville: HRSA; 22 Mar 2023. Archived PDF. Accessed 7 Oct 2026.
33. Health Resources and Services Administration. Newborn Screening for Krabbe Disease: A Summary of the Evidence and Advisory Committee Decision (brief report, 13 Mar 2023) and Executive Summary (Oct 2023). Rockville: HRSA. Archived brief report. Accessed 7 Oct 2026.
34. Kemper AR, Lam KK, Ream M, DiCostanzo K, Grosse SD, Prosser LA, et al. Expedited Evidence-Based Review of Newborn Screening for Krabbe Disease: Final Report, February 1, 2024. Prepared for the Maternal and Child Health Bureau (HRSA). Archived PDF. Accessed 7 Oct 2026.
35. Health Resources and Services Administration. Newborn Screening for Infantile Krabbe Disease: A Summary of the Evidence and Advisory Committee Decision (expedited review report, 1 Feb 2024). Rockville: HRSA. Archived PDF. Accessed 7 Oct 2026.
36. Gnanakumar A, Aljaberi R, Laney DA, Martin C, Keller SR, Parikh SH, et al. Peripheral Neuropathy as an Early Marker in Newborn-Screened Krabbe Disease: The Value of Pre-Confirmatory Neurophysiological Testing. J Peripher Nerv Syst. 2025;30(3):e70040. doi:10.1111/jns.70040. PMID:40560063.
37. Mochel F, Sedel F. 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 2.1–2.3 (tables 2.7 and 2.13). doi:10.1007/978-3-662-63123-2_2.
38. Escolar ML, Poe MD, Martin HR, Kurtzberg J. A staging system for infantile Krabbe disease to predict outcome after unrelated umbilical cord blood transplantation. Pediatrics. 2006;118(3):e879–e889. doi:10.1542/peds.2006-0747. PMID:16923928.
39. Wright MD, Poe MD, DeRenzo A, Haldal S, Escolar ML. Developmental outcomes of cord blood transplantation for Krabbe disease: A 15-year study. Neurology. 2017;89(13):1365–1372. doi:10.1212/WNL.0000000000004418. PMID:28855403.
40. Zhang J, Tan Q, Zhu W, Yi H, Wu Y. Survival benefit and immune-related toxicities after hematopoietic stem cell transplantation in Krabbe disease: a systematic review and meta-analysis. Front Immunol. 2026;17:1924827. doi:10.3389/fimmu.2026.1924827. PMID:42677154.
41. Ardissone A, Ferrera G, Bonanomi S, Farina L, Ciano C, Nucera S, et al. Long-term neurological outcome after hematopoietic stem cell transplant in juvenile Krabbe disease. J Neurol. 2026;273(2):104. doi:10.1007/s00415-025-13607-x. PMID:41604001.
42. Escolar ML, West T, Dallavecchia A, Poe MD, LaPoint K. Clinical management of Krabbe disease. J Neurosci Res. 2016;94(11):1118–1125. doi:10.1002/jnr.23891. PMID:27638597.
43. ClinicalTrials.gov. NCT05739643: A Phase 1/2 Clinical Study of Intravenous AAVrh10 Vector Expressing GALC in Krabbe Subjects Who Previously Received Hematopoietic Stem Cell Transplantation (REKLAIM). Forge Biologics. Registry. Last updated 20 Nov 2025; accessed 6 Oct 2026.
44. ClinicalTrials.gov. NCT06308718: Long-term Follow-up Study to Evaluate Safety and Efficacy of FBX-101 in Krabbe Patients. Forge Biologics. Registry. Terminated; last updated 5 Sep 2025; accessed 6 Oct 2026.
45. European Medicines Agency. EU/3/21/2511: Adeno-associated virus serotype rh10 containing the human GALC gene for the treatment of Krabbe disease [Internet]. Orphan designation. Accessed 6 Oct 2026.
46. ClinicalTrials.gov. NCT04771416: Study of Safety, Tolerability and Efficacy of PBKR03 in Pediatric Subjects With Early Infantile Krabbe Disease. Gemma Biotherapeutics. Registry. Suspended; last updated 30 Jun 2026; accessed 6 Oct 2026.
47. European Medicines Agency. EU/3/21/2417: Adeno-associated virus serotype hu68 containing the human GALC gene for the treatment of Krabbe disease [Internet]. Orphan designation. Accessed 6 Oct 2026.
48. ClinicalTrials.gov. NCT02254863: Augmentation of Umbilical Cord Blood Transplantation for Inherited Metabolic Diseases With Intrathecal Administration of Human Umbilical Cord Blood-Derived Oligodendrocyte-Like Cells (DUOC-01). Registry. Accessed 6 Oct 2026.
49. ClinicalTrials.gov. NCT07740512: Safety, Tolerability, Pharmacokinetics, and Preliminary Efficacy of PLX-200 in Pediatric Patients With Lysosomal Storage Disorders (SOTERIA). Polaryx Therapeutics. Registry. Accessed 6 Oct 2026.
50. Lin DS, Ho CS, Huang YW, Lee TH, Huang ZD, Wang TJ, et al. Durable Global Correction of CNS and PNS and Lifespan Rescue in Murine Globoid Cell Leukodystrophy via AAV9-Mediated Monotherapy. Cells. 2025;14(24):1942. doi:10.3390/cells14241942. PMID:41439962.
51. Nam BG, Seo JH, Hong SA, Kim JH, Kang M, Notario GR, et al. In vivo adenine base editing of mutant Galc gene ameliorates Krabbe disease progression. Genome Med. 2026;18(1):46. doi:10.1186/s13073-026-01620-2. PMID:42026681.
52. Cascino F, Ricca A, Picciotti I, Valeri E, Unali G, Saporito V, et al. Chimeric enzymes enhance treatment potential for globoid cell leukodystrophy through hematopoietic stem cell gene therapy. Mol Ther. 2025;33(12):6226–6247. doi:10.1016/j.ymthe.2025.09.030. PMID:40988335.
53. Firth HV, Hurst JA. Carrier frequency and testing for autosomal recessive disorders. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 811.
54. Barczykowski AL, Foss AH, Duffner PK, Yan L, Carter RL. Death rates in the U.S. due to Krabbe disease and related leukodystrophy and lysosomal storage diseases. Am J Med Genet A. 2012;158A(11):2835–2842. doi:10.1002/ajmg.a.35624. PMID:22991292.
55. ClinicalTrials.gov. NCT02993796: Krabbe Disease Global Patient Registry. State University of New York at Buffalo. Registry. Accessed 6 Oct 2026.
56. Watanabe S, Kimura D, Sasaki T, Yamamoto T, Saito J, Minakawa M. Thoracic surgery for spontaneous pneumothorax in a long-term survivor of infantile Krabbe disease: a case report. J Cardiothorac Surg. 2026;21(1):349. doi:10.1186/s13019-026-04187-6. PMID:42015219.