Keywords
leukodystrophy - myelin - adrenoleukodystrophy - metachromatic leukodystrophy - bone
marrow transplantation - enzymatic replacement therapy - hematopoietic stem cell transplantation
Leukodystrophies are a group of genetically determined disorders that affect development
or maintenance of central nervous system (CNS) myelin.[1]
[2]
[3]
[4] Leukodystrophies have a reported incidence of 1 in 7500 live births, but fewer than
half of patients receive a specific diagnosis.[5] Diagnosis can be based on a combination of history, expected prevalence, physical
and neurologic examination, radiological features, or laboratory findings, including
genetic testing.[5]
[6] Leukodystrophies can present at any age from infancy to adulthood, with variability
in disease progression and clinical presentation ranging from developmental delay
to seizures to spasticity. Although there are few cures, there are significant opportunities
for care and improvements in patient well-being. Their high prevalence, combined with
rapid advances in imaging, genetics, and potential treatments, makes an understanding
of the leukodystrophies necessary for care providers in genetics, bone marrow transplantation,
and neurology.
One confusing issue is the definition of the term “leukodystrophy.” For the purposes
of this review, a leukodystrophy will refer to a presumed genetically determined abnormality
in myelin development, maintenance, or turnover. In contrast, a “leukoencephalopathy”
will be defined more broadly as any abnormality of the white matter, including not
only leukodystrophies,[2] but also abnormalities of the white matter, such as caused by trauma, toxicity,
or insult, for example, periventricular leukomalacia secondary to premature birth
or white matter abnormalities caused by chemotherapy or meningitis.
Myelin Development
Myelin development begins during fetal life and continues through adulthood. It involves
a complex developmental orchestration of genes, proteins, and different cell types.
Production of myelin requires oligodendrocytes, the glial cells that produce myelin
and the myelin sheath, and interactions with other cell types, particularly neurons.
Myelin is necessary for normal action potential propagation and nervous system function,
maintenance of axonal health, and nutritional support of axons.[7]
Myelin is produced by oligodendrocytes (also known as oligodendroglia). Oligodendrocytes
are generated during fetal and postnatal life by oligodendrocyte precursor cells (OPCs).
Oligodendrocyte precursor cells can persist into adult life,[8] and there is evidence that new oligodendrocytes formed during adult life generate
new myelin.[9] Each oligodendrocyte extends many myelin sheet-forming processes that envelope axonal
projections. Different processes from a single oligodendrocyte typically envelope
different axons.[10] Thus, on any axon, different oligodendrocytes can be contributing to neighboring
stretches of myelin. Myelin is composed of proteins (30%) and lipids (70%), and its
production is highly energy-dependent. The most abundant protein components are myelin
basic protein (MBP) and proteolipid protein (PLP). The lipid components of myelin
are cholesterol, phospholipids, and glycolipids; glycosphingolipids, particularly
galactocerebrosides, are the predominant glycolipids.[11]
[12]
In addition to the oligodendrocyte-specific determinants of myelin formation, neuron-oligodendrocyte
interactions are also necessary for normal myelination.[13] Oligodendrocytes must migrate to the correct locations in the brain to effect axonal
ensheathment by myelin,[14] and the extent of myelination is affected by neuronal activity.[15]
During brain development, myelination proceeds in an inside–outside and caudal-to-rostral
fashion, with myelination of deeper structures occurring first. Myelination begins
by the fourth month of gestation. Essentially all areas of the brain are myelinated
by the second year of life, although the total myelin content of the brain does not
peak until the third decade of life.[16]
It is apparent, then, that a large number of insults or processes, during development
or in adulthood, could affect myelin development or the maintenance of myelin, and
hence lead to a leukodystrophy. The biological complexity of myelin development requires
both neurons and glia, and requires their interactions. Therefore, problems with any
of these three could produce a leukodystrophy. In addition, leukodystrophies can arise
from defects in intrinsic/biochemical pathways that are common to all cells, but that
are magnified in glia due to their high energy requirements and extensive membrane
turnover. For example, mutations in the intrinsic myelin protein PLP1 lead to Pelizaeus-Merzbacher disease, and vanishing white matter disease (VWMD) is
caused by mutations in the ubiquitously expressed translation initiation factor subunits
EIF2B1–5 that are required in all cells for protein production.
Despite improvements in our fundamental understanding of myelin development, there
remain numerous confusing aspects to the pathogenesis of many of the leukodystrophies.
Why can VWMD be asymptomatic until initiated by a sudden precipitant?[17] Why can the same mutation in X-linked adrenoleukodystrophy be asymptomatic, or cause
an adult-onset neuropathy, or lead to a fatal childhood cerebral inflammatory demyelination?[18] Research into these topics will lead to improved treatment for patients and significant
insights into myelin biology.
Categories and Diseases
Significant confounders for discussing the different leukodystrophies are several.
First, there is disagreement over the exact definition of a leukodystrophy.[5]
[19] Second, only about half of all causes of leukodystrophies are known.[5]
[20] Finally, limited understanding of the pathogenesis of the leukodystrophies makes
classification difficult.
Currently, leukodystrophies appear to fall into three broad categories ([Table 1]): (1) hypomyelination, in which there is absent or diminished myelin production; (2) dysmyelination, in which there is abnormal myelin development; and (3) demyelination, in which there is loss and/or destruction of previously established myelin. Although
widely accepted (or perhaps tacitly assumed) that demyelinating inflammatory diseases
such as multiple sclerosis or neuromyelitis optica constitute a different diagnostic
category than demyelinating leukodystrophies,[21] it will be interesting in the next decade to see if shared common elements of pathogenesis
are discovered.[22] This being said, trials of immune modulatory therapy for demyelinating leukodystrophies
have not shown efficacy.[23]
Table 1
Categories, disease, and gene name abbreviations of representative leukodystrophies
|
Category
|
Gene
|
Diagnosis
|
MRI Features
|
|
Hypomyelinating
|
|
|
|
|
Pelizaeus-Merzbacher disease
|
PLP1
|
G
|
|
|
Pelizaeus-Merzbacher-like disease
|
GJA12
|
G
|
|
|
4H syndrome
|
POLR3A, POLR3B
|
G
|
Cerebellar atrophy
|
|
Salla disease/sialuria
|
SLC17A5
|
Urine sialic acid
|
|
|
Fucosidosis
|
FUCA1
|
Urine oligosaccharides, LLE
|
|
|
HABC/Other
|
TUBB4A
|
G
|
Atrophy of basal ganglia, cerebellum
|
|
Dysmyelination/demyelination[a]
|
|
|
|
|
Metachromatic leukodystrophy
|
ARSA
|
Urine sulfatides, LLE
|
Frontal onset, symmetric
|
|
MLD-like
|
PSAP
|
Urine sulfatides, G
|
frontal onset, symmetric
|
|
SUMF1
|
Urine mucopolysaccharides
|
Frontal onset, symmetric
|
|
Krabbe (Globoid cell leukodystrophy)
|
GALC1
|
LLE
|
Optic nerve enlargement, DGM
|
|
Canavan disease
|
ASPA
|
Urine NAA, G
|
Macrocephaly, increased NAA on MRS
|
|
Alexander disease
|
GFAP
|
G
|
Frontal onset, megalencephaly, contrast enhancement, BG, T, BS
|
|
Aicardi-Goutieres syndrome
|
TREX1, RNASEH2A, RNASEH2B, RNASEH2C, SAMHD1, ADAR
|
CSF pleocytosis, CSF IFN-α, G
|
Calcifications, temporal cysts
|
|
Megalencephalic leukodystrophy with cysts
|
MLC1, HEPACAM
|
G
|
Megalencephaly, cysts
|
|
X-linked adrenoleukodystrophy
|
ABCD1
|
Plasma VLCFA
|
Contrast Enhancement, parieto-occipital onset
|
|
Mitochondrial diseases
|
various
|
Lactate (serum, CSF), G
|
BG, BS, DGM
|
|
Vanishing white matter disease (VWMD)
|
EIF2B, subunits 1–5
|
G
|
Stranding in white matter, white matter isointense with CSF
|
|
VWMD-like disease
|
LMNB1
|
G
|
Stranding in white matter, white matter isointense with CSF
|
Abbreviations: BG, basal ganglia involvement; BS, brainstem involvement; CSF, cerebrospinal
fluid; DGM, involvement of deep gray matter; G, gene test; LLE, leukocyte lysosomal
enzymes; MRS, magnetic resonance spectroscopy; T, thalamus involvement; VLCFA, serum
very-long chain fatty acids.
Note. “Diagnosis” indicates the primary initial method for diagnosis; “MRI features”
indicate any particularly unique and/or distinguishing features observed on MRI associated
with the disease.
a Dysmyelination and demyelination disorders are included together, as many of these
diseases may have both characteristics at different points in their natural history.
However, certain leukodystrophies do not fit clearly into the above categories. For
example, Rett syndrome and Batten disease can have white matter changes and clinical
features that would initially suggest a leukodystrophy before subsequent testing leads
to their ultimate diagnosis. Or some patients have syndromic dysmorphic physical features,
and are found to have chromosome microdeletions or microduplications, such as in the
18q- syndrome[24]; yet the primary CNS pathology can consist of leukodystrophy.
An alternative method of classification is based on the primary site of action of
the gene product implicated in the leukodystrophy. In this regard, leukodystrophies
such as metachromatic leukodystrophy or Krabbe disease can be linked to defects in
the lysosome, while X-linked adrenoleukodystrophy can be linked to defects in the
peroxisome.[4] A drawback to the site-of-action classification approach is the imperfect correlation
between gene function and pathology of the leukodystrophy. For example, X-linked adrenoleukodystrophy
is caused by mutation of the ABCD1 gene producing a peroxisomal transporter protein, yet levels and presence of very-long
chain fatty acids correlate with neither the development of disease nor the timing
of disease onset.
Clinical Presentation
There are few reliable clinical indicators of leukodystrophies: The symptoms can be
nonspecific, and the age of onset can vary from prenatal to adult. After history and
examination, magnetic resonance imaging (MRI) is the key diagnostic test that should
prompt evaluation for a specific leukodystrophy diagnosis.
A few leukodystrophies can have classic presentations, and a specific diagnosis should
be considered. In a child with acute deterioration of neurologic status, mitochondrial
disease or VWMD should be considered. Infants with Pelizaeus-Merzbacher disease often
present with nystagmus and/or head titubation. Macrocephaly is a common feature in
Canavan disease and Alexander disease, whereas microcephaly is a feature of Aicardi-Goutieres
disease. Skin color changes and/or adrenal insufficiency (e.g., Addison disease) in
a school-age boy should lead to evaluation for X-linked adrenoleukodystrophy.
In neonates and infants, presenting symptoms can include encephalopathy or developmental
delay. In children, teens, and adults, symptoms can be more insidious, ranging from
behavioral or psychiatric changes, loss of formerly achieved milestones or deterioration
in skills, vision changes, ataxia, or gait problems (often from spasticity). Although
not necessarily a presenting symptom, seizures are much more common than previously
realized in children with leukodystrophies, and affect up to 49% of children.[5]
Incidence and Prevalence
The most recent population-based estimate for leukodystrophies shows an incidence
of 1 in 7,663 live births, although this was felt to be an underestimate, with a “true”
incidence closer to 1:7,500.[5] A population-based determination for X-linked adrenoleukodystrophy in Norway revealed
a birth prevalence of 1.6 in 100,000.[25] In Poland, the birth prevalence of metachromatic leukodystrophy has been estimated
at 4.1 in 100,000.[26] To date, there have been few large studies from Africa, the Middle East, or Asia
to examine relative incidences or prevalences of the different leukodystrophies, and
it will be important and interesting to learn the disease distribution from these
regions.
The prevalence of different leukodystrophies has also been difficult to determine;
furthermore, in most studies almost half of all patients remain without a definitive
diagnosis.[5]
[20] In the Utah cohort, the most common diagnoses (in descending order of prevalence)
were metachromatic leukodystrophy, Pelizaeus-Merzbacher disease, mitochondrial diseases,
X-linked adrenoleukodystrophy, and VWMD.[5] Most studies have not compared incidences of different leukodystrophies. For many
years, leukodystrophy-specific incidences were estimated based on case determinations
at tertiary referral centers. Given the referral bias in case ascertainment, these
estimates must be interpreted with caution. That being said, and using a restrictive
definition of leukodystrophy, an incidence of 2 in 100,000 live births was determined
from referral centers in Germany.[27] X-linked adrenoleukodystrophy has been estimated at 1 in 100,000 in France, 1.6
in 100,000 in Australasia, and 1 in 30,000 in Japan.[28]
[29]
[30]
Magnetic Resonance Imaging and Neuroimaging
Magnetic Resonance Imaging and Neuroimaging
Magnetic resonance imaging is the current gold standard for diagnosis of a leukodystrophy.
Although computed tomography (CT) can indicate abnormal signal density in the white
matter, MRI is necessary for the determination of signal characteristics that can
provide potentially diagnostic information. A large amount of careful work has been
performed on the use of MRI and features that can be used for diagnosis.[6]
[31]
[32]
[33] Key features on MRI include the presence of contrast enhancement; the presence of
cysts, calcifications, or more subtle structural abnormalities (such as periventricular
garlands in Alexander disease)[34]; hypomyelination versus high T2-signal abnormality; and the relative T1-signal hyper-
or hypointensity.[6] However, even though MRI algorithms are helpful for diagnosis,[6] they are not able to provide a final diagnosis for many patients because of the
need for skilled radiological interpretation and experience and limitations of their
sensitivity and specificity.[31]
Complications and limitations to the use of MRI include the appearance and characteristics
of myelin change with normal development,[35] and this can be confounding for the interpretation of what is normal or abnormal
on MRI. In particular, until the age of 2 years children are normally relatively hypomyelinated
compared with adults. Further, there is a range (or bell curve) of normal development
of myelin appearance on MRI, and at age 2 years some children will not be fully myelinated.
Also, hypomyelination can be a result or complication of developmental delay or other
systemic disorders or illnesses.[36]
[37] Another complicating feature is that the T1 and T2 signal characteristics of myelin
shift between birth and 1 year; in infants, myelin is bright on T1 and dark on T2.
By 1 year of life, myelin is dark on T1 and bright on T2. Despite caveats, it is important
to broadly recognize leukodystrophies by certain key features on MRI, and to differentiate
these from potential leukodystrophy mimics ([Fig. 1]; [Table 1]).
Fig. 1 Illustrative examples of magnetic resonance images of different leukodystrophies,
and of leukodystrophy mimics. Images are axial slices. (A) T1 Image of a hypomyelinating leukodystrophy, TUBB4A. (B) T1 Image with contrast of a demyelinating leukodystrophy, X-linked adrenoleukodystrophy
(X-ALD). (C) T2 Fluid-attenuated inversion-recovery (FLAIR) image of metachromatic leukodystrophy
(MLD). (D) T2 FLAIR image of vanishing white matter disease (VWMD). (E) T2 FLAIR image of multiple sclerosis (MS). (F) T2 FLAIR image of acute disseminated encephalomyelitis (ADEM).
Diagnosis
How does one diagnose a new patient with a leukodystrophy? This continues to be a
vexing problem for the clinician. High costs of genetic testing, combined with the
likelihood that many causes of leukodystrophy have not yet been discovered, continue
to limit the ability to make diagnoses. There is no universally accepted algorithm,
and no estimation of costs or sensitivity and specificity for different approaches
to diagnosis.
Currently, the most sensible approach appears to combine three elements: (1) testing
for treatable diseases, (2) testing based on MRI features, and (3) using relative
prevalence to guide testing ([Fig. 2]). We propose an algorithm for leukodystrophy diagnosis that is rational and cost
effective, with a tiered approach to testing. There is utility in making an expedient
diagnosis by excluding treatable forms of leukodystrophies, curtailing other expensive
and lengthy testing, and providing valuable reassurance and prognostic information
to the patient and their family.
Fig. 2 A proposed diagnostic algorithm for testing of leukodystrophies. CK, creatine kinase;
MRI, magnetic resonance imaging; PMD, Pelizaeus-Merzbacher disease; VWMD, vanishing
white matter disease.
Steady advances in next-generation sequencing technologies are making whole exome
sequencing a first-tier option for diagnosis of complex genetic disorders, with reported
yields of 25%.[38] Unbiased genome-wide approaches exemplified by whole exome or whole genome sequencing
provide the potential for diagnosis of known diseases without step-wise ordering of
multiple individual tests. Further, genome-wide sequencing can contribute to the ongoing
discovery of novel disease genes.
Next-generation sequencing technologies provide the potential for unbiased diagnosis
of known diseases without individual ordering of multiple individual tests, and will
contribute to the discovery of novel disease genes. However, continued limitations
and problems associated with this technology include (1) substantial cost—up to $15,000
to $20,000—although these numbers are rapidly decreasing; (2) potential for identifying
unanticipated disease variants unrelated to the test indication; (3) potential false-negatives
due to imperfect exome coverage; and (4) methodological limitations in the interpretation
phase if a clear disease-associated variant is not identified. This last problem is
significant due to the large number of deleterious gene variants in all humans that
could plausibly be related to a phenotype (especially in the CNS), which could yield
false-positive associations.
Whole-exome sequencing continues to become more accessible, and may become the method
of choice for the diagnosis of leukodystrophies. A rational, cost-effective algorithm
for leukodystrophy diagnosis could be developed, with a tiered approach to testing.
For example, first-tier testing could include assessment for treatable leukodystrophies:
leukocyte lysosomal enzyme testing for metachromatic leukodystrophy, and in males,
serum very-long chain fatty acids, for adrenoleukodystrophy. The second tier could
be whole-exome sequencing. However, without sufficient data on the sensitivity of
whole-exome sequencing for leukodystrophies, and of the false-negative rate for missing
gene deletions or duplications (such as for Pelizaeus-Merzbacher or VWMD), this approach
is not yet indicated.
Newborn screening for leukodystrophies is being developed in several states.[39]
[40] Screening has been piloted in New York for Krabbe disease,[41]
[42] but is now also being considered in other states, including California and Minnesota.
However, because of variability in disease course, variability in penetrance and disease
progression, as well as limited treatment options, there continues to be debate about
how widely, and for what diseases, newborn screening should be considered.[43]
[44]
[45]
Treatments
Treatment options for leukodystrophies are disappointingly sparse at the current time.
Essentially, the only curative treatment option is bone marrow transplant (BMT)/hematopoietic
stem cell transplantation (HSCT). Bone marrow transplant is only available for a subset
of leukodystrophies, chiefly X-linked adrenoleukodystrophy, metachromatic leukodystrophy,
and Krabbe disease,[46]
[47]
[48] and is only helpful if transplantation is performed prior to substantial disease
progression. Patients who have BMT are at high risk for disease progression; first,
because there is a time lag between transplant and effective “rescue,” and second,
because the BMT process itself appears to accelerate disease progression in some patients.
Bone marrow transplant is not always successful, and carries a substantial mortality
risk, approaching 20%. Donor sources for BMT include cord blood, bone marrow, or peripheral
blood stem cells.[49]
[50]
Enzyme replacement therapy (ERT) is an option for some of the lysosomal disorders
(Gaucher disease; Fabry disease; mucopolysaccharidosis types I, II, and VI; and Pompe
disease), which can have leukodystrophy as a component.[51] Enzyme replacement therapy has shown some efficacy in animal models in other leukodystrophies.[52] Clinical trials are being pursued, but convincing effectiveness has not yet been
shown.
Lorenzo's oil is a treatment that has been proposed to reduce progression of X-linked
adrenoleukodystrophy in its cerebral form.[53] A 4:1 mixture of glyceryl trioleate and glyceryl trierucate, Lorenzo's oil can normalize
levels of very-long chain fatty acids in plasma, although as previously noted, these
do not show a correlation with disease progression, and current published data to
do not demonstrate efficacy for inhibiting disease progression or altering outcomes.[54]
[55]
[56] In the United States, Lorenzo's oil is available only in clinical trials.
Care for Patients
Although we have seen that treatment options for leukodystrophies are limited, there
are tremendous opportunities for improving the care for patients.[57] This patient-centric approach should lead to discussions between the clinician,
patient, and family about what care and treatment is most important and most helpful.
As has been demonstrated for other currently incurable genetic conditions (e.g., cystic
fibrosis), the strategies of routine symptomatic care can have a profound impact on
both the quality and the duration of a patient's life.[58] Nationally and internationally, there are no standardized guidelines for care or
treatment of leukodystrophy patients, which results in wide variability in care and
costs: > sevenfold difference in costs across children's hospitals in the United States.[59]
Complications of disease, even if the disease itself is not progressive, can lead
to progressive disability requiring assistance for mobility and activities of daily
living, and even surgery.[60]
[61] Leukodystrophy patients can have significant health care requirements and costs,
driven largely by in-patient admissions.[5]
[62] A relatively small subset of leukodystrophy patients (15%) have much greater health
care problems, suggesting that potential interventions for these patients could have
a proportionately greater impact. As expected, patients who undergo BMT have much
higher costs. However, even taking into account BMT, patients with infections and
patients needing mechanical ventilation have higher costs and health care needs.[62]
Building from this analysis of health care utilization, a recent study showed that
infection rates in leukodystrophy patients correlate with potentially modifiable risk
factors[63]: for example, failure to vaccinate against seasonal flu significantly increases
the risk for hospitalization with influenza, and urinary tract infections are associated
with the presence of indwelling urinary catheters. Although these are common sense
issues, they also outline a path for potential clinical care guidelines that could
be implemented at this time to reduce hospitalizations and improve care.
Future Directions
The development of new treatments is and should be a primary goal. In this regard,
studies in model organisms are leading to a better understanding of the molecular
underpinnings of the leukodystrophies, which will provide foundations for new therapeutic
options. The best examples include X-linked adrenoleukodystrophy and Alexander disease.
Adrenoleukodystrophy has been studied in mouse knockout models[64]
[65]
[66]
[67] although failing to recapitulate CNS degeneration, mouse knockout models have proven
useful for the testing of anti-inflammatory therapies.[67] Interestingly, a powerful genetic model of adrenoleukodystrophy is the Drosophila
bubblegum mutant, which manifests two key hallmarks of the human disease: age-dependent CNS
degeneration and very-long chain fatty acid accumulation.[68] For Alexander disease, both Drosophila and mice harboring the mutated isoform of
human GFAP (glial fibrillary acidic protein) as a transgene exhibit Rosenthal fiber
formation and neurodegeneration, which has been shown to originate from glia.[69]
[70] Additional models are likely to emerge in the near future. It is also important
to note that although animal models recapitulate key diagnostic aspects of diseases,
they also have the potential to mimic the incomplete penetrance and variable expressivities
that characterize human conditions. Thus, animal models will provide a platform to
explore the environmental and genetic interactions that make each case presentation
so unique.
From the perspective of existing therapies, two significant hurdles to their implementation
are the absence of efficient delivery to the CNS, an organ system largely protected
by the blood–brain barrier, and the lack of robust methodologies to deliver gene replacement
in affected oligodendrocytes or neurons. In addition, significant pathology may happen
early in development, in the first year of life, or even before birth. There is significant
interest in using stem cells or modified induced pluripotent stem cells (iPSCs) for
the leukodystrophies.[71] Although their use has great potential, their practical use still does not seem
imminent. Gene therapy is also under active investigation, but with only a few examples
and considerable technical challenges, it remains experimental.[72]
[73] A limitation of some of the gene therapy approaches is that they use BMT to provide
the source of the enzyme, which thus introduces all of the potential morbidities of
BMT. Novel drug discovery, or repurposing of known drugs, is another promising avenue
of current therapies for several leukodystrophies.[56]
[67]
[74]
A corollary to the need for improved treatments is the need for faster, definitive
diagnosis. Whole exome or whole genome sequencing appears to be on the threshold of
becoming the diagnostic method of choice, as it can avoid the “diagnostic odyssey”
faced by many patients.[75]
[76] Sequencing is as cost effective as a brain MRI, and with improved data analysis
will become as rapid. However, there are unanswered questions about the disconnect
between the underlying gene or biochemical defect, as in X-linked adrenoleukodystrophy
or Krabbe disease, and actually developing the disease.[39]
[77] Further, unresolved ethical questions surround the implementation of whole genome
sequencing as a screening or general diagnostic tool.[78]
[79]
Conclusions
The field of knowledge encompassing leukodystrophies is rapidly expanding. Within
this decade new methods of diagnosis, such as next-generation sequencing, as well
as new therapies will revolutionize patient care. Though exciting, we can still improve
our care and improve the lives of patients today by increased attention to preventable
and modifiable features of the leukodystrophy diseases.
