What Is Metachromatic Leukodystrophy?

Metachromatic leukodystrophy (MLD) is a rare, inherited disease in which the body cannot break down certain fats called sulfatides, leading to progressive destruction of the protective myelin coating around nerve fibers in both the brain and the peripheral nervous system. It is caused by a deficiency of the enzyme arylsulfatase A (ARSA), and without functional myelin, children and adults with MLD gradually lose the ability to walk, talk, swallow, and think.¹ The disease has gained renewed attention in recent years because a gene therapy, atidarsagene autotemcel, became the first treatment approved specifically for MLD in several countries, though it works best when given before symptoms appear.

What Happens Inside the Body

MLD belongs to a family of conditions known as lysosomal storage disorders. Every cell has tiny compartments called lysosomes that act as recycling centers, breaking down worn-out molecules. In MLD, the enzyme ARSA is missing or barely functional, so sulfatides cannot be broken down properly. They pile up in cells that produce myelin, the insulating sheath that allows electrical signals to travel quickly along nerves. As sulfatide accumulates, myelin deteriorates in a process called demyelination, which affects both the white matter of the brain and the nerves running to the limbs and organs.

Research in laboratory models has shown that sulfatide buildup triggers a cascade of toxic effects on cells, including overactivation of a DNA-repair enzyme called PARP-1. In one study using brain-tissue slice cultures treated with sulfatides, PARP-1 expression in surviving cells nearly doubled compared with controls, and blocking PARP-1 with the drug olaparib significantly reduced that toxic response.² Findings like these are still early-stage, but they help explain why demyelination in MLD is so aggressive and why stopping it requires more than simply slowing sulfatide production.

The Different Forms and How They Present

MLD is not a single clinical picture. It is divided into subtypes based on the age when symptoms first appear, and the subtype heavily influences how fast the disease progresses and which symptoms come first.

Late-infantile (onset by about age 2.5): The most common and most severe form. Children typically develop normally at first, then begin losing motor milestones. Gait disturbances and abnormal movement patterns are usually the earliest signs, appearing at a median age of about 1.5 years. Language acquisition may stall; roughly half of these children never learn to speak in complete sentences, and those who do often lose expressive language entirely by about age three.³
Early-juvenile (onset roughly age 2.5 to 6): Progression is similarly rapid. Motor symptoms dominate at the start, with loss of ambulation and onset of seizures resembling the late-infantile pattern, though loss of trunk control and swallowing difficulties tend to occur somewhat later.⁴
Late-juvenile (onset age 6 to 16): Problems with concentration, behavior, and fine motor skills often appear alongside gait disturbances. Cognitive decline unfolds over years rather than months, with reading, writing, and math skills eroding about four years after onset.⁵
Adult-onset (age 16 or older): The rarest and slowest-progressing form. Cognitive and psychiatric symptoms predominate, with mild or absent peripheral neuropathy and lower brain-imaging severity at plateau.⁶

One critical finding from a large nationwide study is that the type of first symptom matters as much as the age of onset. In late-juvenile and adult patients, those who first showed motor symptoms progressed just as rapidly as children with early-onset disease. But patients who presented with only cognitive symptoms at onset had significantly milder disease courses, regardless of age.⁷ This makes the first symptom a surprisingly powerful predictor of how the disease will unfold.

Peripheral Neuropathy as an Early Clue

Because MLD attacks myelin throughout the body, the peripheral nerves in the arms and legs are affected alongside the brain. Polyneuropathy is essentially universal in the late-infantile form and common in all other subtypes.⁸ In infants, slowed nerve conduction can sometimes be detected before other symptoms are obvious, with conduction velocities dropping to about 30 percent of what is normal for age.⁹ This peripheral nerve involvement is one reason why neurologists sometimes pick up MLD before brain imaging shows much change, and it also explains the early loss of reflexes and the floppy-limbed appearance that parents may notice.

Why Diagnosis Is Tricky

Getting the time from first symptoms to confirmed diagnosis down is a persistent challenge. In the late-infantile form, the gap averages about a year; in the juvenile form, it stretches to about two years.¹⁰ Several factors conspire to delay recognition.

Adult-onset MLD is particularly prone to misdiagnosis. Because it progresses slowly and begins with cognitive or psychiatric symptoms, it is often mistaken for early-onset dementia or schizophrenia, sometimes for years before the correct diagnosis is reached.¹¹

Even when MLD is suspected and enzyme testing is ordered, interpreting the results can be complicated. A common pitfall is the ARSA pseudodeficiency allele, a harmless genetic variant carried by a substantial fraction of the general population that causes low ARSA enzyme activity in lab tests without causing disease. A person who carries one true MLD mutation and one pseudodeficiency allele can have very low enzyme levels, creating a false alarm. The key to telling these individuals apart from actual MLD patients is measuring sulfatide in the urine: true MLD patients excrete far more sulfatide than pseudodeficiency carriers.¹²

Another diagnostic wrinkle emerged from a case involving a patient from Kenya whose ARSA activity was low but not dramatically so under standard lab conditions. When the substrate concentration used in the test was lowered, the patient’s enzyme activity fell to about 10 percent of normal, a level consistent with MLD. The standard assay had partly masked the deficiency because high substrate concentrations can artificially compensate for certain missense mutations.¹³ Cases like this underscore why genetic sequencing has become an important complement to enzyme testing.

The Saposin B Variant

Most MLD is caused by mutations in the ARSA gene, but a small number of cases result from a deficiency of saposin B, a helper protein needed for ARSA to reach and break down sulfatides properly. In saposin B-deficient MLD, ARSA enzyme levels test within the normal range, which can throw clinicians off entirely. Diagnosis requires genetic testing of the prosaposin (PSAP) gene.¹⁴ Saposin B deficiency can present in both childhood and adulthood,¹⁵ and because so few mutations have been reported, clinical experience with this form remains limited. Anyone with a clinical picture consistent with MLD but normal ARSA activity should be evaluated for saposin B deficiency.

Newborn Screening Efforts

Because every available treatment for MLD works best when started before symptoms appear, there is strong interest in adding MLD to routine newborn screening panels. A pilot study screened over 27,000 newborn dried blood spots using a two-tier strategy: first measuring sulfatide levels by mass spectrometry, then checking ARSA activity in any samples above the cutoff. Out of the full cohort, 195 newborns initially flagged as having elevated sulfatides, but all except two turned out to be false positives once ARSA activity was measured. Of those two, genetic sequencing confirmed one as an affected MLD patient and the other as a carrier.¹⁶

The results demonstrated that mass spectrometry-based sulfatide screening is feasible, but also highlighted the challenge of a high initial false-positive rate that must be filtered by a second-tier enzyme test.¹⁷ Several countries and U.S. states are now evaluating MLD newborn screening programs, driven by the availability of gene therapy that has a narrow intervention window. Consensus guidelines in the United States emphasize that standardized presymptomatic monitoring is critical once an at-risk newborn is identified, because the time between a positive screen and the optimal treatment window may be very short.¹⁸

Hematopoietic Stem Cell Transplantation

Before gene therapy arrived, the main disease-modifying option was a bone marrow or cord blood transplant, formally called hematopoietic stem cell transplantation (HSCT). The idea is to replace the patient’s faulty bone marrow with donor cells that produce normal ARSA. Over time, some donor-derived cells migrate into the brain and become microglia, the brain’s resident immune cells, which can then release functional enzyme to surrounding tissue.

HSCT can slow disease progression, but results have been mixed. In the largest single-institution report, transplanted patients were less likely to lose gross motor or language function than untransplanted patients, and brain MRI severity scores were significantly lower at the latest examination.¹⁹ However, transplant-related mortality was significant: four of 24 transplanted children died from complications of the procedure itself. Among the survivors, about half showed stable disease, while the other half continued to decline. The patients most likely to benefit were those transplanted early, before major motor or cognitive deficits set in, with an IQ of at least 85 and relatively low MRI severity scores.

Another large cohort study reached a similar conclusion: later-onset phenotypes stand the best chance when transplant happens pre-symptomatically, but HSCT remains limited and should be weighed carefully on a case-by-case basis.²⁰ The fundamental problem with HSCT is timing and potency: it takes months for donor cells to engraft and begin producing enzyme in the brain, and the amount of enzyme they deliver may not be enough to keep up with the rapid myelin destruction in early-onset forms.

Gene Therapy With Atidarsagene Autotemcel

Gene therapy for MLD takes the transplant concept a step further. Instead of relying on a matched donor, clinicians collect a patient’s own blood stem cells, use a lentiviral vector to insert working copies of the ARSA gene, and then infuse the corrected cells back after chemotherapy conditioning. Because the inserted gene can be engineered to produce enzyme at higher-than-normal levels, the corrected microglia that eventually settle in the brain can flood surrounding tissue with far more ARSA than a standard transplant provides.

Mouse studies established the proof of concept: gene-corrected microglia secreted active ARSA that was taken up by neurons throughout the brain, including Purkinje cells in the cerebellum, where up to 80 to 90 percent of those neurons showed detectable enzyme.²¹ In human trials, the results have been striking. An integrated analysis of clinical trial and expanded-access data published in The Lancet found that most treated patients progressively acquired motor skills within the range expected of healthy children or stabilized their motor performance, with maintained ability to walk. Most also showed normal cognitive development and prevention or delay of demyelination throughout follow-up. The benefits were most pronounced in patients treated before symptoms appeared.²²

One lingering concern has been whether gene therapy can protect the peripheral nerves, not just the brain. A recent study looking specifically at nerve conduction in treated late-infantile patients found that by 36 months of age, treated children showed nerve conduction velocities roughly 15 meters per second higher than untreated age-matched controls across both motor and sensory nerves, a meaningful improvement in the severity of demyelinating neuropathy.²³ Peripheral neuropathy was already present in most treated patients at their pre-treatment examination, which underscores how early myelin damage begins and why presymptomatic treatment matters.

Research is also advancing on next-generation lentiviral vectors. One improved vector design produced roughly 27-fold more ARSA than the current clinical vector in laboratory microglia and also loaded more functional enzyme into small secreted vesicles that can shuttle ARSA to neighboring cells.²⁴ If this translates to humans, future gene therapies could potentially be effective at lower doses, reducing manufacturing complexity and possibly improving safety.

Experimental Approaches Beyond Gene Therapy

For patients who are diagnosed too late for gene therapy or HSCT to be effective, enzyme replacement delivered directly into the spinal fluid is being explored. Intrathecal recombinant ARSA involves repeated injections into the cerebrospinal fluid. In a study of children treated with this approach, high-dose intrathecal ARSA appeared to slow the loss of brain gray matter volume and delay demyelination compared with untreated children or those receiving lower doses. Gray-matter volume in treated children correlated strongly with motor function scores over time.²⁵ The treatment requires repeated spinal injections, though, which is burdensome. Researchers are developing implantable cell-based devices that could continuously release ARSA into the spinal fluid, potentially replacing the need for repeated injections.²⁶

Animal work has also shown that timing is everything for any intervention. When gene therapy using a viral vector was given to mice that already had neurological symptoms, the treatment reduced sulfatide accumulation in certain brain regions but did not reverse behavioral deficits, likely because irreversible damage had already occurred.²⁷ This finding reinforces the urgency around early diagnosis and presymptomatic treatment.

The Cost and Access Challenge

Atidarsagene autotemcel (sold as Libmeldy) is one of the most expensive medicines ever approved, with a list price in the millions. Cost-effectiveness analyses for the U.S. market have estimated that the therapy provides gains in excess of 23 quality-adjusted life years for presymptomatic patients compared with best supportive care, and that it falls below common willingness-to-pay thresholds from both payer and societal perspectives.²⁸ In other words, when the price is weighed against the decades of severe disability it can prevent, the therapy arguably pays for itself in the long run. Still, the upfront sticker shock is real for health systems, and different countries are experimenting with outcome-based payment models in which the drug company is reimbursed only if the patient meets certain clinical milestones over time. Modeling in the Netherlands found that outcome-based agreements reduce financial risk for payers and can be considerably more favorable than simple negotiated discounts, especially if a patient does not respond as well as predicted.²⁹

What Families Actually Go Through

The clinical literature on MLD focuses heavily on enzyme levels and MRI scores, but the reality for families is grinding and all-consuming. An international survey of MLD caregivers found that the average family logged about 30 outpatient hospital visits and four inpatient stays per year. Half of caregivers said their child’s diagnosis had negatively affected their relationship with a spouse or partner, and more than three-quarters had either stopped working entirely or shifted to part-time employment. Rates of anxiety and depression among caregivers exceeded population norms.³⁰ A separate survey of U.K. and Irish caregivers echoed these findings, with employment, finances, relationships, and personal health all affected.³¹

Supportive care for symptomatic patients includes management of spasticity, seizures, feeding difficulties, and pain, ideally coordinated by specialists familiar with MLD who can follow the patient over time.³² Physical therapy, speech therapy, nutritional support, and assistive devices become increasingly important as the disease advances. For families navigating this, connecting with MLD-specific patient organizations can provide both practical resources and emotional support that the medical system alone does not offer.