Neurology · Medical Literature
Central nervous system histopathological findings in classic infantile Pompe disease: a systematic review with clinical relevance.
Abstract
Classic infantile Pompe disease is the most severe phenotype within the clinical spectrum of Pompe disease and is characterized by central nervous system (CNS) involvement. A systematic review on histopathological findings in classic infantile Pompe disease was conducted using Embase, Medline Ovid, Web of science and Cochrane central databases for original research articles until 16-12-2025. Articles reporting a histological examination of the CNS were selected. Classic infantile Pompe disease was defined by hypertrophic cardiomyopathy, symptom onset < 12 months of age, and GAA deficiency, if available. Of the 3734 records identified, 39 articles were included in the study, comprising 49 patients, 7 of whom treated with enzyme replacement therapy. The median age at autopsy was 6 months (range 1.2-17.0) for untreated and 12 months (range 7.5-21.0) for treated patients. Histologic abnormalities were found throughout the CNS. The cerebral and cerebellar white matter (WM), globus pallidus, dentate nucleus, motor nuclei of the brainstem, and anterior horn cells were most frequently affected. Glycogen accumulation was severe in astrocytes and in neurons of the globus pallidus. Cortical neurons were affected with greater variability, and subcortical structures were more frequently involved than cortical ones. Understanding of CNS involvement in classic infantile Pompe disease is rapidly evolving. Histopathological studies from children up to 21 months demonstrate early and widespread CNS pathology. Notably, there is a marked temporal gap between the extensive abnormalities observed at autopsy in infancy and later radiological, biochemical, and clinical manifestations. While white matter pathology is already prominent in young patients, abnormalities on magnetic resonance imaging typically emerge from 2 to 3 years of age, biomarker changes from around 5 years, and cognitive decline, reduced processing speed, and background activity slowing on electroencephalogram from approximately 8-10 years onward. These observations suggest a gradual evolution of CNS disease, although direct clinicopathological correlations remain limited. Epilepsy has been reported later in the disease course, and we have observed upper motor neuron signs in older patients, suggesting progressive gray and white matter involvement that may contribute to cognitive decline. The mechanisms linking glycogen accumulation to neurological manifestations remain incompletely understood. Taken together, these observations form the basis for several hypotheses on the pathogenesis of CNS involvement in classic infantile Pompe disease.
Clinical relevance
Classic infantile Pompe disease (CIPD), the most severe phenotype of Pompe disease, is characterized by early and widespread CNS pathology that is increasingly recognized as a critical determinant of long-term neurological outcome, even in enzyme replacement therapy (ERT)-treated survivors. This systematic review addresses a significant knowledge gap by consolidating histopathological evidence of CNS involvement across 49 autopsy cases, underscoring that CNS disease begins well before it is detectable by conventional neuroimaging or clinical assessment. As ERT extends patient survival, clinicians must anticipate a progressive CNS disease trajectory that manifests decades after the initial histopathological insult, making early surveillance and neuroprotective strategy development urgent clinical priorities.
Key findings
1. Of 3,734 records screened, 39 articles met inclusion criteria, comprising 49 patients (42 untreated, 7 ERT-treated), making this the most comprehensive synthesis of CNS histopathology in CIPD to date. 2. Median age at autopsy was 6 months (range 1.2–17.0) in untreated patients and 12 months (range 7.5–21.0) in ERT-treated patients, confirming that severe CNS pathology is present in early infancy. 3. Histological abnormalities were distributed throughout the CNS; the most frequently affected regions were the cerebral and cerebellar white matter, globus pallidus, dentate nucleus, motor nuclei of the brainstem, and anterior horn cells. 4. Glycogen accumulation was most severe in astrocytes and in neurons of the globus pallidus; cortical neuronal involvement was present but showed greater inter-case variability, and subcortical structures were more consistently affected than cortical ones. 5. A marked temporal dissociation exists between early histopathological CNS disease (present by age 6–12 months at autopsy) and later clinical manifestations: MRI white matter abnormalities typically emerge at age 2–3 years, biomarker changes around age 5 years, and cognitive decline, reduced processing speed, and EEG background slowing from approximately 8–10 years onward. 6. Progressive gray and white matter involvement is further evidenced by the later emergence of epilepsy and upper motor neuron signs in older surviving patients, suggesting ongoing neurodegeneration beyond the infantile period; however, direct clinicopathological correlations remain limited by study design and case ascertainment.
Practice implications
Clinicians managing ERT-treated CIPD survivors should implement structured longitudinal CNS surveillance protocols—including serial brain MRI, neuropsychological testing, and EEG—beginning by age 2–3 years, even in patients who appear neurologically intact, given the significant lag between subclinical histopathological disease and detectable clinical or radiological manifestations. Patient selection for emerging CNS-targeted therapies (e.g., substrate reduction, next-generation ERT formulations with CNS penetrance, or gene therapy) should prioritize presymptomatic or early-symptomatic individuals, as the histopathological data suggest a window of opportunity before irreversible neuronal loss and white matter degeneration accrue. Multidisciplinary teams should counsel families that cognitive decline, processing speed deficits, epilepsy, and upper motor neuron signs represent expected disease milestones in long-term survivors, and that current ERT does not adequately address CNS glycogen clearance, highlighting an unmet therapeutic need that should guide trial design and outcome measure selection.