Advertisement

Evaluating brain white matter hyperintensity, IQ scores, and plasma neurofilament light chain concentration in early-treated patients with infantile-onset Pompe disease

  • Yu-Kang Hsu
    Affiliations
    Department of Pediatrics, Taipei City Hospital Renai Branch, Taipei, Taiwan
    Search for articles by this author
  • Yin-Hsiu Chien
    Affiliations
    Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan

    Department of Pediatrics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Department of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan
    Search for articles by this author
  • Steven Shinn-Forng Peng
    Affiliations
    Department of Medical Images, National Taiwan University Hospital, Taipei, Taiwan
    Search for articles by this author
  • Wuh-Liang Hwu
    Affiliations
    Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan

    Department of Pediatrics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Department of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan
    Search for articles by this author
  • Wang-Tso Lee
    Affiliations
    Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan

    Department of Pediatrics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Department of Pediatric Neurology, National Taiwan University Children’s Hospital, Taipei, Taiwan
    Search for articles by this author
  • Ni-Chung Lee
    Affiliations
    Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan

    Department of Pediatrics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Department of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan
    Search for articles by this author
  • Eric Po-Yu Huang
    Affiliations
    Department of Radiology, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
    Search for articles by this author
  • Wen-Chin Weng
    Correspondence
    Correspondence and requests for materials should be addressed to Wen-Chin Weng, Department of Pediatrics, National Taiwan University Hospital, No. 8, Chung-Shan South Road, Taipei, Taiwan 100
    Affiliations
    Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan

    Department of Pediatrics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Department of Pediatric Neurology, National Taiwan University Children’s Hospital, Taipei, Taiwan
    Search for articles by this author
Open AccessPublished:November 17, 2022DOI:https://doi.org/10.1016/j.gim.2022.10.005

      ABSTRACT

      Purpose

      The study aimed to describe central nervous system (CNS) progression in patients with infantile-onset Pompe disease (IOPD) and explore the potential clinical impact and predictors.

      Methods

      Patients with IOPD treated with enzyme replacement therapy were longitudinally followed with brain magnetic resonance imaging (MRI) and evaluation for IQ scores from 2004 to 2021. Investigation of CNS involvement focused on white matter (WM) abnormalities and was quantified using a scoring system for metachromatic leukodystrophy. MRI scores were correlated with plasma neurofilament light chain (NfL) concentration and IQ scores.

      Results

      A total of 19 patients who started enzyme replacement therapy at a mean age of 26 days were analyzed; the median age at last examination was 12.1 (range = 1.7-19) years. MRI abnormalities were found in all patients, from supratentorial central WM to U-fibers, then to infratentorial WM, and eventually to gray matter. MRI scores progressed (n = 16) at variable rates (range = 0.8-2.7/y) and were positively correlated with age (n = 16) and negatively correlated with IQ scores (n = 8). Plasma NfL concentration was positively correlated with MRI scores (r2 = 0.8569; P < .001; n = 13).

      Conclusion

      Our results suggest that the progression of CNS involvement in IOPD may be associated with neuroaxonal injury and decreased IQ scores. NfL could serve as a biomarker for CNS involvement in IOPD.

      Keywords

      Introduction

      Pompe disease (OMIM 232300) is a lysosomal storage disorder caused by the deficiency of GAA, leading to an accumulation of glycogen mainly within the skeletal and cardiac muscles.
      • Hirschhorm R.
      • Reuser A.J.
      Glycogen storage disease type II: acid alpha-glucosidase (acid maltase) deficiency.
      The clinical presentations of Pompe disease are diverse, ranging from rapidly progressive infantile-onset Pompe disease (IOPD) to slowly progressive late-onset Pompe disease. Classic IOPD is characterized by progressive generalized muscle weakness, hypertrophic cardiomyopathy, and respiratory failure before the age of 6 months. If left untreated, patients die in the first year of life because of heart failure.
      • van den Hout H.M.P.
      • Hop W.
      • van Diggelen O.P.
      • et al.
      The natural course of infantile Pompe’s disease: 20 original cases compared with 133 cases from the literature.
      Enzyme replacement therapy (ERT) with recombinant human GAA (rhGAA) can significantly improve patient survival and cardiac outcomes.
      • Kishnani P.S.
      • Corzo D.
      • Nicolino M.
      • et al.
      Recombinant human acid [alpha]-glucosidase: major clinical benefits in infantile-onset Pompe disease.
      ,
      • Chen L.R.
      • Chen C.A.
      • Chiu S.N.
      • et al.
      Reversal of cardiac dysfunction after enzyme replacement in patients with infantile-onset Pompe disease.
      Initiating ERT as soon as possible through the early detection of IOPD in newborn screening programs can further lead to significant motor improvements and a better quality of life and prognosis.
      • Chien Y.H.
      • Hwu W.L.
      • Lee N.C.
      Pompe disease: early diagnosis and early treatment make a difference.
      ,
      • Chien Y.H.
      • Lee N.C.
      • Thurberg B.L.
      • et al.
      Pompe disease in infants: improving the prognosis by newborn screening and early treatment.
      However, ERT cannot cross the blood-brain barrier, and there is increasing evidence of central nervous system (CNS) involvement concurrent with prolonged survival in patients with IOPD.
      • Korlimarla A.
      • Lim J.A.
      • Kishnani P.S.
      • Sun B.
      An emerging phenotype of central nervous system involvement in Pompe disease: from bench to bedside and beyond.
      • Chien Y.H.
      • Lee N.C.
      • Peng S.F.
      • Hwu W.L.
      Brain development in infantile-onset Pompe disease treated by enzyme replacement therapy.
      • Chien Y.H.
      • Lee N.C.
      • Chen C.A.
      • et al.
      Long-term prognosis of patients with infantile-onset Pompe disease diagnosed by newborn screening and treated since birth.
      Autopsy studies of patients with IOPD have shown glycogen accumulation in the neurons and glial cells of the cerebral cortex, basal ganglia, brainstem, cerebellum, and spinal cord.
      • Thurberg B.L.
      • Lynch Maloney C.
      • Vaccaro C.
      • et al.
      Characterization of pre- and post-treatment pathology after enzyme replacement therapy for Pompe disease.
      ,
      • Pena L.D.M.
      • Proia A.D.
      • Kishnani P.S.
      Postmortem findings and clinical correlates in individuals with infantile-onset Pompe disease.
      Neuroimaging studies with brain magnetic resonance imaging (MRI) have also reported structural abnormalities, including white matter (WM) abnormalities, transient ventricular enlargement, and extra-axial cerebrospinal fluid accumulation in patients with IOPD who are treated with rhGAA.
      • Chien Y.H.
      • Lee N.C.
      • Chen C.A.
      • et al.
      Long-term prognosis of patients with infantile-onset Pompe disease diagnosed by newborn screening and treated since birth.
      ,
      • Messinger Y.H.
      • Mendelsohn N.J.
      • Rhead W.
      • et al.
      Successful immune tolerance induction to enzyme replacement therapy in CRIM-negative infantile Pompe disease.
      • McIntosh P.T.
      • Hobson-Webb L.D.
      • Kazi Z.B.
      • et al.
      Neuroimaging findings in infantile Pompe patients treated with enzyme replacement therapy.
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      The severity of WM abnormalities appears to increase with age
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      ; however, few studies have explored the evolving pattern of WM involvement, its association with long-term cognitive function, and potential prognostic biomarkers correlated with the pathogenesis of WM abnormalities in patients with IOPD.
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      • Ebbink B.J.
      • Poelman E.
      • Plug I.
      • et al.
      Cognitive decline in classic infantile Pompe disease: an underacknowledged challenge.
      We recently showed axonal degeneration and loss in aged GAA knockout Pompe mice.
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      In the past decade, neurofilament light chain (NfL) protein levels in the cerebrospinal fluid and blood have been shown to be a sensitive but unspecific marker of axonal injury and neuronal damage.
      • Gaetani L.
      • Blennow K.
      • Calabresi P.
      • Di Filippo M.
      • Parnetti L.
      • Zetterberg H.
      Neurofilament light chain as a biomarker in neurological disorders.
      Neurofilaments, including NfL, medium chain, and heavy chain, are the major cytoskeleton proteins in myelinated axons, and axonal injury and neuronal damage can lead to the release of these proteins into the cerebrospinal fluid or blood.
      • Perrot R.
      • Berges R.
      • Bocquet A.
      • Eyer J.
      Review of the multiple aspects of neurofilament functions, and their possible contribution to neurodegeneration.
      In recent years, sensitive measurement techniques have made it possible to measure NfL in blood samples, in which the concentration is approximately 40-fold lower than that in the cerebrospinal fluid.
      • Gaetani L.
      • Blennow K.
      • Calabresi P.
      • Di Filippo M.
      • Parnetti L.
      • Zetterberg H.
      Neurofilament light chain as a biomarker in neurological disorders.
      ,
      • Disanto G.
      • Barro C.
      • Benkert P.
      • et al.
      Serum neurofilament light: a biomarker of neuronal damage in multiple sclerosis.
      NfL levels in serum or plasma, measured using these novel methods, have been verified to be correlated with cerebrospinal fluid NfL levels in a variety of CNS disorders with neuroaxonal injury, including multiple sclerosis, traumatic brain injury, stroke, autosomal dominant Alzheimer disease, and metachromatic leukodystrophy (MLD).
      • Gaetani L.
      • Blennow K.
      • Calabresi P.
      • Di Filippo M.
      • Parnetti L.
      • Zetterberg H.
      Neurofilament light chain as a biomarker in neurological disorders.
      ,
      • Disanto G.
      • Barro C.
      • Benkert P.
      • et al.
      Serum neurofilament light: a biomarker of neuronal damage in multiple sclerosis.
      • Beerepoot S.
      • Heijst H.
      • Roos B.
      • et al.
      Neurofilament light chain and glial fibrillary acidic protein levels in metachromatic leukodystrophy.
      • Sánchez-Valle R.
      • Heslegrave A.
      • Foiani M.S.
      • et al.
      Serum neurofilament light levels correlate with severity measures and neurodegeneration markers in autosomal dominant Alzheimer’s disease.
      Axonal degeneration of unhealthy neurons has been shown to result in WM abnormalities in a GAA knockout mouse model,
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      and swollen neurons with glycogen accumulation have been found in postmortem findings in patients with IOPD.
      • Pena L.D.M.
      • Proia A.D.
      • Kishnani P.S.
      Postmortem findings and clinical correlates in individuals with infantile-onset Pompe disease.
      Therefore, we hypothesized that plasma NfL, an indicator of axonal injury, could serve as a potential biomarker to predict the severity of WM abnormalities in patients with IOPD. To test this hypothesis, in this prospective study, we explored the natural history of brain WM abnormalities in patients with IOPD using a scoring system previously developed to evaluate MLD.
      • Eichler F.
      • Grodd W.
      • Grant E.
      • et al.
      Metachromatic leukodystrophy: a scoring system for brain MR imaging observations.
      We also analyzed plasma NfL levels and cognitive function in patients with IOPD to investigate whether plasma NfL levels could reflect the severity of WM abnormalities and the effect of CNS involvement on cognitive outcomes.

      Materials and Methods

      Participants

      This cohort study was conducted from 2004 to 2021 and included children with confirmed IOPD from National Taiwan University Hospital. The diagnosis of Pompe disease was established according to severe deficiency of GAA activity and the presence of 2 GAA disease-associated variants in trans, using the same methodology as that reported in our previous studies.
      • Chien Y.H.
      • Lee N.C.
      • Chen C.A.
      • et al.
      Long-term prognosis of patients with infantile-onset Pompe disease diagnosed by newborn screening and treated since birth.
      ,
      • Chien Y.H.
      • Tsai W.H.
      • Chang C.L.
      • et al.
      Earlier and higher dosing of alglucosidase alfa improve outcomes in patients with infantile-onset Pompe disease: evidence from real-world experiences.
      All of the included children presented with symptoms by age 6 months and were diagnosed as having IOPD. In addition, all of the included children were positive for cross-reactive immunologic material and were treated with rhGAA (alglucosidase alpha, Myozyme [Genzyme Corporation]) with a dosage determined according to their clinical condition.
      • Chien Y.H.
      • Tsai W.H.
      • Chang C.L.
      • et al.
      Earlier and higher dosing of alglucosidase alfa improve outcomes in patients with infantile-onset Pompe disease: evidence from real-world experiences.
      Patients who died or did not undergo follow-up brain MRI at our hospital were excluded. The clinical information, including the clinical outcomes with regard to the occurrence of motor decline and the biochemical outcomes with regard to serum creatine kinase and urine glucose tetrasaccharide levels, of patients 1 to 18 have been described previously.
      • Chien Y.H.
      • Tsai W.H.
      • Chang C.L.
      • et al.
      Earlier and higher dosing of alglucosidase alfa improve outcomes in patients with infantile-onset Pompe disease: evidence from real-world experiences.
      Demographic data, including genetic variants, cardiac function, motor development, and ventilation status were acquired from medical records. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

      Neuroimaging acquisition and analysis

      Brain MRI was performed using a standard birdcage head coil on a 1.5-T Sonata MRI system (Siemens). All scannings included a minimum of T1-weighted, T2-weighted, and fluid-attenuated inversion recovery images with a slice thickness of approximately 4 to 5 mm. In our previous study, we found abnormal brain development in patients with IOPD treated with ERT. Therefore, in this study, we conducted prospective and longitudinal brain MRI examinations in all patients regardless of CNS symptoms.
      • Chien Y.H.
      • Lee N.C.
      • Peng S.F.
      • Hwu W.L.
      Brain development in infantile-onset Pompe disease treated by enzyme replacement therapy.
      The patients underwent the first MRI scan at a mean age of 3.8 (range = 1.5 to 30) months, and the longitudinal brain MRI follow-up was arranged every 1 to 3 years. Sedation was performed if the patient did not cooperate well during the MRI examination. Particular attention was paid to the presence or absence of WM abnormalities and the extent of involvement. Although all images were evaluated, brain WM abnormalities were quantified on the T2-weighted and fluid-attenuated inversion recovery images using a scoring system previously developed to evaluate MLD. We chose this MLD scoring system because we found a similar pattern of WM involvement in both MLD and IOPD, which was also mentioned in another study.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      This scoring system yields a severity score ranging from 0 to 34 points according to the location and extent of WM involvement and the presence of atrophy.
      • Eichler F.
      • Grodd W.
      • Grant E.
      • et al.
      Metachromatic leukodystrophy: a scoring system for brain MR imaging observations.
      In brief, the major areas assessed were the supratentorial WM, frontopontine-corticospinal projection fibers (corpus callosum, internal capsule, and infratentorial WM, including midline pons and cerebellar WM), and deep gray nuclei (thalamus, basal ganglia). Both cerebellum and cerebral atrophy were also scored. The supratentorial WM regions were subcategorized into the frontal, anterior temporal, and parieto-occipital WM, and each region was further subdivided into periventricular, central, and subcortical U-fibers. The anterior and posterior limbs of the internal capsule were evaluated separately, and the genu and splenium of the corpus callosum were also evaluated separately. The severity of T2 hyperintensity was graded as 1 point (faint hyperintensity) or 2 points (dense hyperintensity). A total of 89 separate MRI scans from 19 patients were reviewed and scored by 2 independent pediatric neurologists (Y.-K.H. and W.-C.W.). If their opinions diverged, a consensus was made after discussion with a board-certified neuroradiologist (S.S.-F.P), whose opinion was considered final.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      To further understand the brain involvement pattern over time, we divided brain regions according to the MLD scoring system and defined the age at an event as at least 1 point in the brain region on the MLD scoring system. We then estimated the time-to-event using Kaplan-Meier analysis to evaluate the age at involvement in different brain regions across all 89 MRI scans.
      To evaluate the effect of the severity of CNS involvement on cognitive outcomes, we assessed the correlation between brain MRI scores and IQ scores. We also analyzed the correlation between plasm NfL levels and brain MRI scores to investigate whether plasma NfL levels could reflect the severity of WM abnormalities.
      During the study period, we obtained exploratory images of 2 patients to better illustrate the findings on brain MRI. Another 2 patients further underwent concomitant diffusion tensor imaging with the imaging parameters described previously.
      • Peng S.S.F.
      • Chien Y.H.
      • Hwu W.L.
      • Liu H.M.
      • Tseng I.W.Y.
      Cerebral diffusion tensor images in infants and neonates with infantile onset Pompe disease.
      Fractional anisotropy (FA), which assesses the degree of anisotropic diffusion occurring within a region, was determined as previously described.
      • Tseng C.H.
      • Chien Y.H.
      • Lee N.C.
      • et al.
      Gene therapy improves brain white matter in aromatic L-amino acid decarboxylase deficiency.
      FA maps of the whole brain were obtained from each subject. One patient also underwent brain computed tomography (CT) owing to unprovoked seizures, which manifested as generalized tonic-clonic seizures for 2 minutes with hypoxia and then resolved after 2 doses of intravenous lorazepam as part of clinical care to exclude acute symptomatic etiologies.

      Cognitive assessment

      Cognitive function was assessed in children aged >3 years using an age-appropriate standardized IQ test administered by a clinical child psychologist. The Wechsler Preschool and Primary Scale of Intelligence, revised edition, was administered to children aged 3 to 5 years, and the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) was administered to children aged >6 years. Both full-scale IQ scores and composite scores of the WISC-IV, including verbal comprehension, perceptional reasoning, working memory, and processing speed, were recorded and analyzed.

      Plasma sampling and NfL measurements

      Plasma NfL levels were measured in 17 patients during the last follow-up and 15 age- and sex-matched healthy controls. Plasma samples were collected, centrifuged at 2000g for 10 minutes at room temperature, divided into aliquots, and frozen at –80 °C until analysis. Plasma NfL concentrations were measured using a highly sensitive single-molecule array assay platform (UmanDiagnostics) combined with a Simoa NF-light Advantage Kit (Quanterix Inc). Detailed instructions can be found in a previous study.
      • Kuhle J.
      • Barro C.
      • Andreasson U.
      • et al.
      Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa.
      Batch-prepared calibrators (bovine lyophilized NfL) ranging from 0 to 10,000 pg/mL were stored at –80 °C. The limit of detection (mean blank signal + 3 SD) for the Simoa NfL assay was 0.97 pg/mL, and the lower limit of quantification (mean blank signal + 10 SD) was 2.93 pg/mL when compensated for a 4-fold sample dilution. Samples at a 1:4 dilution were measured in duplicates by blinded technicians. The average repeatability coefficient of variation of a sample with a mean concentration of 4.14 pg/mL was 6.3%, and the interassay coefficient of variation was 5.1%. For a sample with a mean concentration of 178.77 pg/mL, the intra- and interassay coefficients of variation were 0.9% and 4.0%, respectively.

      Statistical analysis

      Descriptive statistics were used to describe continuous and categorical variables, including plasma NfL concentration. Correlations between plasma NfL concentration and MRI scores were analyzed using Pearson correlation coefficients, and simple linear regression analysis was used for bivariate factor analysis. Other data analyses were performed using an independent t test. Survival probabilities in various regions of brain involvement were estimated using the Kaplan-Meier method. All statistical analyses were carried out from July to August 2021, using IBM SPSS Statistics for Windows, version 24 (IBM Corp), and Prism 5 for Windows, version 5.0 (GraphPad Software Inc). A 2-sided P value of <.05 was considered to be statistically significant.

      Results

      Patient demographics

      A total of 19 patients (median [range] age at last examination = 12.1 [1.7-19] years; 11 [57.9%] males) were included in this study, and their clinical characteristics are summarized in Supplemental Table 1. Of the 19 patients, 4 (21.1%) were identified after a clinical diagnosis, and 15 (78.9%) were identified through newborn screening programs (Supplemental Figure 1), including 14 with left ventricular hypertrophy diagnosed before age 1 month and 1 patient (patient 11) before age 3 months. The median (range) age of the patients at ERT initiation was 26 (5-180) days. A total of 8 patients (42.1%) had respiratory impairment at the last follow-up, whereas 6 (31.6%) required noninvasive ventilation, and 2 (10.5%) required invasive ventilation through a tracheostomy.

      WM and functional changes

      T2-weighted WM hyperintensity is the typical abnormality found on MRI of the brain in patients with IOPD. For example, in patient 3, such changes were seen on axial cuts over the pons, periventricular WM, corona radiata, and centrum semiovale, and they were more extensive over the frontal and parieto-occipital regions (Figure 1A-D). In this patient, and possibly other patients, hyperdense lesions (56-58 Hounsfield [H] units) were found on CT images at the corresponding locations (Figure 1E-H). The patient had unprovoked seizures with epileptiform discharges on electroencephalography at age 12.9 years, but there were no signs of CNS infection. WM abnormalities were associated with axonal disorganization, as shown in diffusion tensor imaging analysis (Figure 1I-L). For example, patient 5 had minimal WM changes, and the corresponding color-coded FA maps showed organized WM tracts, whereas patient 8 had extensive WM changes, and the corresponding FA maps showed reduced anisotropic diffusion, implying a lesser degree of WM organization.
      Figure thumbnail gr1
      Figure 1Brain image changes in the patients with infantile-onset Pompe disease. A-D. Axial T2-weighted magnetic resonance imaging images of patient 3 showed hyperintensities (arrowheads) at the levels of the pons (A), periventricular white matter (WM) (B), corona radiata (C), and centrum semiovale (D). E-H. Corresponding computed tomography images also showed abnormal hyperdense lesions (arrowheads). I-L. Axial T2-hyperintensity was correlated with fractional anisotropy (FA) maps. T2-weighted imaging (I) of patient 5 showed minimal WM changes, and the corresponding color-coded FA maps (J) showed organized WM tracts. T2-weighted imaging (K) of patient 8 showed extensive WM changes, and the corresponding FA maps (L) showed disorganized WM tracts with reduced anisotropic diffusion.

      WM abnormality progression with age and genotype

      In total, 16 of 19 (84.2%) patients (patients 1-16) who had at least 3 brain MRI scans during the follow-up were analyzed further (Supplemental Figure 1). Their mean age at the last examination was 12.0 (3.1-19.0) years, and all of them had WM abnormalities during the follow-up, with MRI scans performed every 1 to 3 years. A total of 86 MRI scans were scored, and the mean and median (range) latest MRI scores were 15.9 and 13.5 (7-30), respectively. We found that MRI scores increased with age in all patients (Figure 2), with linear regression model r2 values ranging from 0.9334 to 0.9975 (maximum P = .03) except for patients 9, 15, and 16, all of whom had only 3 MRI scans. In addition, patients 15 and 16 were younger than others at the last MRI scan (aged 3.8 and 3.7 years, respectively). The GAA gene variant p.D645E is a founder gene sequence variation in Chinese patients with Pompe disease. Among the 16 patients, 4 had the p.D645E homozygous sequence variation, 5 had p.D645E plus another allele as nonsense/splicing variations, 4 had p.D645 plus another allele as deletion variations, and 3 had other missense sequence variation combinations. We found that the rate of MRI score progression varied; the patients with p.D645E plus another allele as deletion variations seemed to progress at the fastest rate, whereas 2 patients with other missense variations (patients 5 and 11) progressed at the slowest rate. The maximum rate of MRI score progression was approximately 2.7 points per year (patient 14), whereas the minimum rate was approximately 0.8 points per year (patient 5).
      Figure thumbnail gr2
      Figure 2Trend of an increase in MRI score with age and genotype. Longitudinal MRI scores in 16 patients with infantile-onset Pompe disease are shown. MRI scores increased with age in all patients. The patients are also grouped by genotype: missense + missense (blue), D645E + D645E (brown), D645E + nonsense/splicing (red), and D645E + inframe deletion (gray). Two patients with compound heterozygous missense variants had the slowest rate of increase, whereas 3 patients with D645E + inframe deletion seemed to have the fastest rate. MRI, magnetic resonance imaging; No.1, patient 1, etc.

      Progression of brain MRI abnormalities

      The involvement of WM in the patients revealed a characteristic pattern. The central WM was first involved at a median age of 3.1 (1.5-9.2) years, which was much earlier than the internal capsule (posterior limb), corpus callosum, and infratentorial WM changes found at median ages of 9.1, 11.5, and 11.4 years, respectively (Figure 3A). There was also an anterior-posterior difference in central WM involvement, which was found at a median age of 3.1 years for the parieto-occipital region, 3.6 years for the frontal region, and 6.2 years for the temporal regions (Figure 3B). U-fibers were involved at median ages of 11.4, 11.4, and 14.3 years for the frontal, parieto-occipital, and temporal WM regions, respectively (Figure 3B). Gray matter was involved to a lesser extent than the WM and involved the basal ganglion in 6 patients (37.5%). A total of 6 patients (37.5%) had cerebral atrophy as suggested by ventricular enlargement or inner cerebrospinal fluid space widening at age 8.1 to 11.9 years. Among these 6 patients, 2 (33.3%) advanced to outer cerebrospinal fluid space widening at ages 12.7 and 14.4 years. Patient 1, who had the highest MRI score of 30 points, was noted to have cerebral atrophy at age 11.4 years, basal ganglion involvement at age 11.9 years, and cerebellar atrophy and thalamus involvement at age 14.4 years.
      Figure thumbnail gr3
      Figure 3Kaplan-Meier analysis of MRI changes in the patients. An event was defined as an MRI score of at least 1 point. A. The whole-brain WM was divided into supratentorial WM (red), internal capsule (brown), corpus callosum (blue), and infratentorial WM (brainstem and cerebellum, black), and the median ages at an event were 3.1, 9.1, 11.5, and 11.4 years, respectively. The shaded areas indicate 95% CIs. B. Supratentorial WM, represented in this figure by central WM and U-fibers, was divided into F, P, and T regions. The median ages at involvement of central WM were 3.6, 3.1, and 6.2 years for F, P, and T regions, respectively. The median ages at involvement of U-fibers were 11.4, 11.4, and 14.3 years for F, P, and T regions, respectively. F, frontal; P, parieto-occipital; T, temporal; WM, white matter.

      MRI scores correlated with IQ score and age

      Full-scale IQ scores at the first examination, available in 11 patients (patients 2, 3, 5, 6, and 8-14) at a median age of 4.6 (range = 3.6-9.6, mean = 5.2, SD = 1.9) years, ranged from low to high average (69-117, median = 80) (Figure 4A). Among them, 8 patients (patients 3, 5, 6, 8, 9, and 12-14) had serial IQ measurements. The last IQ scores, at a median age of 10.1 (8.5-13.5) years, ranged from 61 to 118 (median = 73). Two patients had an obvious decline in IQ with age and MRI score, including patient 8 from 69 to 61 and patient 12 from 89 to 70 (Figure 4A), even with appropriate hearing or visual aids or adequate respiratory support to avoid other interfering factors while undergoing the IQ tests.
      Figure thumbnail gr4
      Figure 4Correlation between IQ score and age or MRI score. A. Longitudinal follow-up of full-scale IQ scores in 8 patients (patients 3, 5, 6, 8, 9, and 12-14) and single-point IQ scores in another 3 patients. Except for patient 5, all other patients had a decrease in IQ score with age. B. Correlation between full-scale IQ and MRI scores showed r2 = 0.7331, P = .0077. Regarding the correlation between Wechsler Intelligence Scale for Children-Fourth Edition subdomain score and MRI score, verbal comprehension score showed r2 = 0.5065 and P = .047 (C), working memory score showed r2 = 0.4847 and P = .054 (D), perceptual reasoning score showed r2 = 0.6851 and P = .011 (E), and processing speed score showed r2 = 0.6711 and P = .013 (F). The trend was calculated using linear regression modeling, and the 2 dashed gray lines show 95% CI. MRI, magnetic resonance imaging; No.1, patient 1, etc.
      Full-scale IQ scores also correlated with the most recent MRI scores (r2 = 0.7331; P = .007; n = 8) (Figure 4B). Because the measurement of IQ in patients with IOPD may be complicated by motor function affecting speech and performance, we correlated the WISC-IV subdomain scores with the MRI scores. We found that the subdomains involving minimal motor function, including perceptual reasoning (r2 = 0.6851; P = .01) and processing speed (r2 = 0.6711; P = .01), had correlation with the MRI scores.

      Correlations between plasma NfL concentration and MRI score

      Plasma NfL concentration was measured in 17 patients (median age [range] = 9.97 [0.75-18.09] years) and 15 age- and sex-matched healthy controls (9.05 [0.75-17.3] years). The patients had significantly higher plasma NfL levels (median [range] level = 9.29 [2.31-27.38] pg/mL) than the controls (3.97 [2.40-7.10] pg/mL) (P = .007). In addition, the plasma NfL concentration in the patients varied widely. Of the 17 patients, 8 had similar normal plasma NfL levels as the controls and relatively low MRI scores. Plasma NfL concentrations strongly correlated with MRI scores (r2 = 0.8569; P < .001; n = 13) (Figure 5).
      Figure thumbnail gr5
      Figure 5Correlation between plasma NfL concentration and MRI score. Plasma NfL concentration was higher in the patients with infantile-onset Pompe disease than in the age-matched healthy controls (P = .007). Plasma NfL concentration was also correlated with MRI score (r2 = 0.8569, P < .001). The horizontal dashed line represents the upper limit of the healthy controls. The trend was calculated using linear regression modeling, and the 2 dashed gray lines show the 95% CIs. MRI, magnetic resonance imaging; NfL, neurofilament light chain; No.1, patient 1, etc.

      Discussion

      Progression of MRI abnormalities in the patients with IOPD who were treated early

      There is growing evidence from both animal and human Pompe disease studies of an abnormal accumulation of glycogen in neurons and glial cells due to GAA deficiency.
      • Thurberg B.L.
      • Lynch Maloney C.
      • Vaccaro C.
      • et al.
      Characterization of pre- and post-treatment pathology after enzyme replacement therapy for Pompe disease.
      ,
      • Pena L.D.M.
      • Proia A.D.
      • Kishnani P.S.
      Postmortem findings and clinical correlates in individuals with infantile-onset Pompe disease.
      ,
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      ,
      • Sidman R.L.
      • Taksir T.
      • Fidler J.
      • et al.
      Temporal neuropathologic and behavioral phenotype of 6neo/6neo Pompe disease mice.
      Because the infused GAA protein from ERT does not cross the blood-brain barrier, CNS pathology is likely to progress with ERT. Previous neuroimaging studies have reported WM involvement in patients with IOPD as early as age 18 months.
      • Korlimarla A.
      • Lim J.A.
      • Kishnani P.S.
      • Sun B.
      An emerging phenotype of central nervous system involvement in Pompe disease: from bench to bedside and beyond.
      • Chien Y.H.
      • Lee N.C.
      • Peng S.F.
      • Hwu W.L.
      Brain development in infantile-onset Pompe disease treated by enzyme replacement therapy.
      • Chien Y.H.
      • Lee N.C.
      • Chen C.A.
      • et al.
      Long-term prognosis of patients with infantile-onset Pompe disease diagnosed by newborn screening and treated since birth.
      ,
      • McIntosh P.T.
      • Hobson-Webb L.D.
      • Kazi Z.B.
      • et al.
      Neuroimaging findings in infantile Pompe patients treated with enzyme replacement therapy.
      In this longitudinal study, we showed that these WM abnormalities progressed with age. Although the early institution of ERT extended the life span of the patients (survival to adulthood), brain MRI abnormalities were obviously refractory to ERT.
      In this study, we observed a characteristic pattern of WM involvement, which evolved from supratentorial WM to infratentorial WM and eventually brainstem and cerebellum, which is consistent with a previous study by Ebbink et al.
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      In addition, we confirmed the late involvement of U-fibers, whereas Ebbink et al
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      described sparing of U-fibers in patients with IOPD. In ours and another previous study,
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      although few patients presented with U-fiber involvement at a younger age, abnormalities in the periventricular and central WM still preceded the U-fiber involvement. Moreover, of 14 patients older than 8 years, 6 patients (43%) developed cerebral atrophy between ages 8.1 and 11.9 years, which may be a consequence of WM damage and loss combined with subsequent primary neuronal loss due to glycogen accumulation. Novel volumetric neuroimaging studies that further quantify gray matter and WM volumes may help to elucidate the causes of cerebral atrophy.

      MRI score predicted IQ

      To the best of our knowledge, this longitudinal study contains the largest cohort to date, and thus, we were able to describe systemically the evolving rate of WM changes in each patient using an image scoring system designed initially for MLD. Other scales such as the Fazekas scale have been used to describe the extent and location of CNS involvement
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      ; however, in the Fazekas scale, only 10 anatomical areas of WM including 4 superficial WM and 6 deep WM areas are used. In the MLD scoring system, the superficial WM is subdivided into 9 regions for detailed analysis. Moreover, the MLD score also evaluates deep WM, deep gray matter, cerebellar, and cerebral atrophy, and thus, it may be more comprehensive than the Fazekas scale score to assess CNS abnormalities, including gray matter in the brain. Based on our findings of (1) the strong correlation between MRI scores and patient age, (2) the correlation with IQ scores, and (3) the strong correlation with plasma NfL concentration, we suggest that the MLD scoring system could potentially be applied to patients with Pompe disease and serve as a reliable measurement in longitudinal follow-up studies. Our findings are consistent with the clinical impression of serial MRI scans, other studies on cognitive decline,
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      • Ebbink B.J.
      • Poelman E.
      • Plug I.
      • et al.
      Cognitive decline in classic infantile Pompe disease: an underacknowledged challenge.
      and the hypothetical Pompe axonal damage theory.
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      In subdomain IQ analysis, we also found that the MRI scores were significantly correlated with processing speed and perceptional reasoning. In addition, 2 patients with higher MRI scores (patients 8 and 12) showed a decline in IQ performance, providing clues that the progression of CNS abnormalities may influence the changes in intelligence performance.
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      • Ebbink B.J.
      • Poelman E.
      • Plug I.
      • et al.
      Cognitive decline in classic infantile Pompe disease: an underacknowledged challenge.
      A more precise method to understand WM hyperintensities may be to use a semiautomatic or automatic segmentation clustering algorithm
      • Clas P.
      • Groeschel S.
      • Wilke M.
      A semi-automatic algorithm for determining the demyelination load in metachromatic leukodystrophy.
      ,
      • Tubi M.A.
      • Feingold F.W.
      • Kothapalli D.
      • et al.
      White matter hyperintensities and their relationship to cognition: effects of segmentation algorithm.
      or gray matter/WM volumetry, and we wait for new techniques to be validated in the future.
      Our results showed that WM hyperintensity in brain MRI tends to progress at variable rates, and may be related to the severity of the genotype. Although virtually no enzyme activity is assumed in patients with IOPD, there are still some variations in the degree of residual enzyme activity.
      • Chien Y.H.
      • Lee N.C.
      • Thurberg B.L.
      • et al.
      Pompe disease in infants: improving the prognosis by newborn screening and early treatment.
      Patients with less severe missense variants (patients 5 and 11) had slower increase in MRI scores than those with severe nonsense, insertion, or deletion variants. Two previous studies reported a rapid progression of WM abnormalities within 6 and 12 months in 2 patients with IOPD, and they both had severe genotypes including nonsense and insertion variants.
      • Broomfield A.
      • Fletcher J.
      • Hensman P.
      • et al.
      Rapidly progressive white matter involvement in early childhood: the expanding phenotype of infantile onset Pompe?.
      ,
      • Rohrbach M.
      • Klein A.
      • Köhli-Wiesner A.
      • et al.
      CRIM-negative infantile Pompe disease: 42-month treatment outcome.
      Taken together, these findings may support that different genotypes may affect the rate of WM abnormality progression. However, further studies with more cases are needed to clarify the role of genotypes in CNS progression in IOPD.

      Elevation of plasma NfL indicated neuronal damage

      We found that the increase in plasma NfL concentration in the patients correlated with the MRI score. In addition, the patients with low MRI scores, representing minimal CNS abnormalities, had normal plasma NfL levels as the controls, whereas the patients with high MRI scores had elevated NfL levels. NfL is a sensitive marker of axonal injury and neuronal damage,
      • Perrot R.
      • Berges R.
      • Bocquet A.
      • Eyer J.
      Review of the multiple aspects of neurofilament functions, and their possible contribution to neurodegeneration.
      and elevated NfL levels in the blood have been associated with MRI findings in multiple sclerosis, traumatic brain injury, Alzheimer disease, and MLD.
      • Gaetani L.
      • Blennow K.
      • Calabresi P.
      • Di Filippo M.
      • Parnetti L.
      • Zetterberg H.
      Neurofilament light chain as a biomarker in neurological disorders.
      ,
      • Beerepoot S.
      • Heijst H.
      • Roos B.
      • et al.
      Neurofilament light chain and glial fibrillary acidic protein levels in metachromatic leukodystrophy.
      ,
      • Sánchez-Valle R.
      • Heslegrave A.
      • Foiani M.S.
      • et al.
      Serum neurofilament light levels correlate with severity measures and neurodegeneration markers in autosomal dominant Alzheimer’s disease.
      ,
      • Zetterberg H.
      • Svenningsson A.
      Serum neurofilament light and prediction of multiple sclerosis in clinically isolated syndrome.
      ,
      • Alirezaei Z.
      • Pourhanifeh M.H.
      • Borran S.
      • Nejati M.
      • Mirzaei H.
      • Hamblin M.R.
      Neurofilament light chain as a biomarker, and correlation with magnetic resonance imaging in diagnosis of CNS-related disorders.
      We previously showed axonal destruction and loss in mice with Pompe disease.
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      The NfL levels appear to be not as sensitive as MRI scores in patients with early CNS involvement. In fact, the extent to which plasma NfL levels are elevated correlates with the speed and extent of neuroaxonal damage. In a previous study of patients with MLD,
      • Beerepoot S.
      • Heijst H.
      • Roos B.
      • et al.
      Neurofilament light chain and glial fibrillary acidic protein levels in metachromatic leukodystrophy.
      the plasma NfL level was higher in patients with rapid disease progression than in patients with slow disease progression. Therefore, normal serum NfL levels in patients of IOPD with early CNS involvement may imply the nature of lesions at early stage. That is, the early CNS lesions in patients with IOPD may be mainly related to glycogen accumulation, and the neuroaxonal damage in early CNS involvement may be minimal. Further studies are needed to understand the exact pathophysiology of the evolution of these CNS lesions.
      Peripheral nervous system (PNS) disorders involving neuroaxonal damage can also result in elevated NfL levels. Previous pathologic studies of patients with IOPD have shown glycogen accumulation in Schwann cells and ganglion cells of the PNS
      • Pena L.D.M.
      • Proia A.D.
      • Kishnani P.S.
      Postmortem findings and clinical correlates in individuals with infantile-onset Pompe disease.
      ; however, this finding has rarely been evaluated clinically. In our previous electrophysiological study, we reported that several patients with IOPD showed reduced amplitudes of compound muscle action potential only in the peroneal nerves with absent or impersistent F waves, normal conduction velocity of motor nerves, and normal sensory studies.
      • Tsai L.K.
      • Hwu W.L.
      • Lee N.C.
      • Huang P.H.
      • Chien Y.H.
      Clinical features of Pompe disease with motor neuronopathy.
      These findings suggest that patients with IOPD have mild motor neuronopathy. However, these findings are not present in all patients with IOPD, and our patients had far less PNS involvement than CNS abnormalities. Therefore, we postulate that the effect of PNS involvement on elevated NfL levels is mild. Further large and longitudinal cohort studies of PNS involvement in IOPD are warranted to clarify its clinical significance. In conclusion, NfL may serve as a biomarker to predict CNS involvement in patients with IOPD because myopathy and motor neuropathy
      • Tsai L.K.
      • Hwu W.L.
      • Lee N.C.
      • Huang P.H.
      • Chien Y.H.
      Clinical features of Pompe disease with motor neuronopathy.
      can obscure symptoms of upper motor neuron diseases such as spasticity and hyperreflexia.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      ,
      • Broomfield A.
      • Fletcher J.
      • Hensman P.
      • et al.
      Rapidly progressive white matter involvement in early childhood: the expanding phenotype of infantile onset Pompe?.
      This should be the subject of future research trials.

      Neurologic outcomes of the patients with IOPD

      It is difficult to evaluate IQ in patients with IOPD because IQ measurements can be complicated by motor problems with speech and hand performance. Intelligence can be impaired by involvement of the frontal-parietal integration network.
      • Filley C.M.
      White matter: organization and functional relevance.
      Pompe disease involves the accumulation of glycogen and probably also some degree of postinjury calcium deposition as seen in our CT images. In our study, periventricular and central WM of frontal and parietal regions were typically damaged as a result of glycogen accumulation, which may be related to the decline of intelligence performance. U-fibers are the last part of the brain to myelinate, and they are affected at a later stage of CNS degeneration unless oligodendrocytes are involved.
      • Riley K.J.
      • O’Neill D.P.
      • Kralik S.F.
      Subcortical U-fibers: signposts to the diagnosis of white matter disease.
      In patients with early onset and patients with late onset, the U-fiber remains the last part to be involved in this study as well as other studies,
      • Messinger Y.H.
      • Mendelsohn N.J.
      • Rhead W.
      • et al.
      Successful immune tolerance induction to enzyme replacement therapy in CRIM-negative infantile Pompe disease.
      ,
      • Ebbink B.J.
      • Poelman E.
      • Aarsen F.K.
      • et al.
      Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
      • Korlimarla A.
      • Spiridigliozzi G.A.
      • Crisp K.
      • et al.
      Novel approaches to quantify CNS involvement in children with Pompe disease.
      • Ebbink B.J.
      • Poelman E.
      • Plug I.
      • et al.
      Cognitive decline in classic infantile Pompe disease: an underacknowledged challenge.
      ,
      • Broomfield A.
      • Fletcher J.
      • Hensman P.
      • et al.
      Rapidly progressive white matter involvement in early childhood: the expanding phenotype of infantile onset Pompe?.
      which is also compatible with our previous mouse study with late and less evident glycogen accumulation in oligodendrocytes.
      • Lee N.C.
      • Peng W.H.
      • Tsai L.K.
      • et al.
      Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
      Because we identified neurologic deficits and elevated levels of NfL in this study, further longitudinal studies are needed to evaluate NfL over time and investigate correlations with IQ levels to better understand whether patients are at risk of further deterioration in IQ.

      Limitations

      There were several limitations to this study. First, some patients (especially young patients) refused or failed to cooperate when undergoing MRI examinations, and therefore, the age and extent of WM involvement may have been underestimated. In addition, for gray matter involvement, we examined only deep nuclei owing to technical difficulties. Future studies involving assessments of other gray matter regions will require parameters such as detailed brain volume measurement. Second, only 8 patients had serial IQ data, and we only used the Wechsler Preschool and Primary Scale of Intelligence, revised edition, and WISC-IV to assess patients’ cognition, which may not have been sensitive enough to show the cognitive changes in IOPD. Third, we only had data of single plasma NfL measurements, and follow-up measurements would have strengthened our findings. In addition, elevated plasma NfL levels could be associated with acute brain damage or chronic neuroaxonal damage in both CNS and PNS. However, our patients did not have a history of acute brain damage, and therefore, we believe that the elevation in NfL levels was mostly due to Pompe disease. Further studies evaluating PNS damage in patients with IOPD could help to clarify the contribution of PNS involvement to the elevation in NfL levels. Finally, the relatively homogeneous variants in our cohort may limit the application to other patients with IOPD. However, we believe that this is still a general finding because other studies have reported similar CNS abnormalities. More cases are needed to clarify whether different variants affect the rate of WM hyperintensity progression in brain MRI.

      Conclusion

      In this cohort study, we characterized the CNS abnormalities, focusing on WM, in patients with IOPD. WM abnormalities were found initially in the supratentorial frontal/parietal regions in patients aged as young as 3 years. This was followed by spread to the corpus callosum, internal capsule, gray matter, and infratentorial area with consequent brain atrophy in older patients. We also found that WM involvement occurred and progressed in all patients with IOPD at different rates, with a correlation to plasma NfL levels and decline in cognitive function. Further large-scale and longitudinal studies are warranted. Our study highlights the urgent need to develop therapies targeting neurodegeneration in Pompe disease. Forthermore, neuroimaging quantification tools and plasma NfL concentration may be a promising biomarker for monitoring CNS progression and patient response to treatment in future trials.

      Data Availability

      All data sets used in this study are included in this article. The corresponding author can share deidentified data upon request from a qualified investigator.

      Conflict of Interest

      Y.-H.C. has served on the advisory boards of Amicus Therapeutics, Inc and Sanofi Genzyme, undertaken contracted research for Sanofi Genzyme, and received honoraria and consulting fees from Sanofi Genzyme. W.-L.H. has served on the advisory boards of Audentes Therapeutics and Sanofi Genzyme, undertaken contracted research for Sanofi Genzyme, and received honoraria and consulting fees from Sanofi Genzyme. N.-C.L. has served on the registry advisory boards of Sanofi Genzyme and received honoraria from Sanofi Genzyme. All other authors declare no conflicts of interest.

      Acknowledgments

      Sanofi/Genzyme partially funded the long-term follow-up study (GZ-2014-11166). None of the authors received any honorarium, grant, or other form of payment to produce the manuscript. The study sponsor had no involvement in the study design; collection, analysis, and interpretation of the data; writing of the report; and decision to submit the manuscript for publication.

      Author Information

      Conceptualization: Y.-K.H., Y.-H.C., W.-C.W.; Data Curation: Y.-K.H., W.-C.W.; Formal Analysis: Y.-K.H., Y.-H.C., W.-C.W.; Investigation: Y.-K.H., Y.-H.C., W.-C.W.; Methodology: Y.-H.C., W.-C.W., S.S.-F.P.; Resources: Y.-H.C., W.-C.W., W.-L.H., W.-T.L., N.-C.L.; Supervision: W.-L.H., W.-T.L.; Visualization: S.S.-F.P.; Writing-original draft: Y.-K.H., W.-C.W.; Writing-review and editing: W.-C.W., Y.-H.C., W.-L.H.

      Ethics Declaration

      The Institutional Review Board of National Taiwan University Hospital approved this longitudinal cohort study (National Taiwan University Hospital-Research Ethics Committee No: 200703045R, ClinicalTrials.gov: NCT02399748). Written informed consent was obtained from the children’s parents.

      Additional Information

      References

        • Hirschhorm R.
        • Reuser A.J.
        Glycogen storage disease type II: acid alpha-glucosidase (acid maltase) deficiency.
        in: Scriver C.R. Beaudet A.L. Sly W.S. The Metabolic and Molecular Bases of Inherited Disease. The McGraw-Hill Companies, 2001: 3389-3420
        • van den Hout H.M.P.
        • Hop W.
        • van Diggelen O.P.
        • et al.
        The natural course of infantile Pompe’s disease: 20 original cases compared with 133 cases from the literature.
        Pediatrics. 2003; 112: 332-340https://doi.org/10.1542/peds.112.2.332
        • Kishnani P.S.
        • Corzo D.
        • Nicolino M.
        • et al.
        Recombinant human acid [alpha]-glucosidase: major clinical benefits in infantile-onset Pompe disease.
        Neurology. 2007; 68 (Published correction appears in Neurology. 2008;71(21):1748. https://doi.org/10.1212/01.wnl.0000251268.41188.04): 99-109
        • Chen L.R.
        • Chen C.A.
        • Chiu S.N.
        • et al.
        Reversal of cardiac dysfunction after enzyme replacement in patients with infantile-onset Pompe disease.
        J Pediatr. 2009; 155: 271-275.e2https://doi.org/10.1016/j.jpeds.2009.03.015
        • Chien Y.H.
        • Hwu W.L.
        • Lee N.C.
        Pompe disease: early diagnosis and early treatment make a difference.
        Pediatr Neonatol. 2013; 54: 219-227https://doi.org/10.1016/j.pedneo.2013.03.009
        • Chien Y.H.
        • Lee N.C.
        • Thurberg B.L.
        • et al.
        Pompe disease in infants: improving the prognosis by newborn screening and early treatment.
        Pediatrics. 2009; 124: e1116-e1125https://doi.org/10.1542/peds.2008-3667
        • Korlimarla A.
        • Lim J.A.
        • Kishnani P.S.
        • Sun B.
        An emerging phenotype of central nervous system involvement in Pompe disease: from bench to bedside and beyond.
        Ann Transl Med. 2019; 7: 289https://doi.org/10.21037/atm.2019.04.49
        • Chien Y.H.
        • Lee N.C.
        • Peng S.F.
        • Hwu W.L.
        Brain development in infantile-onset Pompe disease treated by enzyme replacement therapy.
        Pediatr Res. 2006; 60: 349-352https://doi.org/10.1203/01.pdr.0000233014.84318.4e
        • Chien Y.H.
        • Lee N.C.
        • Chen C.A.
        • et al.
        Long-term prognosis of patients with infantile-onset Pompe disease diagnosed by newborn screening and treated since birth.
        J Pediatr. 2015; 166: 985-991.e1https://doi.org/10.1016/j.jpeds.2014.10.068
        • Thurberg B.L.
        • Lynch Maloney C.
        • Vaccaro C.
        • et al.
        Characterization of pre- and post-treatment pathology after enzyme replacement therapy for Pompe disease.
        Lab Invest. 2006; 86: 1208-1220https://doi.org/10.1038/labinvest.3700484
        • Pena L.D.M.
        • Proia A.D.
        • Kishnani P.S.
        Postmortem findings and clinical correlates in individuals with infantile-onset Pompe disease.
        JIMD Rep. 2015; 23: 45-54https://doi.org/10.1007/8904_2015_426
        • Messinger Y.H.
        • Mendelsohn N.J.
        • Rhead W.
        • et al.
        Successful immune tolerance induction to enzyme replacement therapy in CRIM-negative infantile Pompe disease.
        Genet Med. 2012; 14: 135-142https://doi.org/10.1038/gim.2011.4
        • McIntosh P.T.
        • Hobson-Webb L.D.
        • Kazi Z.B.
        • et al.
        Neuroimaging findings in infantile Pompe patients treated with enzyme replacement therapy.
        Mol Genet Metab. 2018; 123: 85-91https://doi.org/10.1016/j.ymgme.2017.10.005
        • Ebbink B.J.
        • Poelman E.
        • Aarsen F.K.
        • et al.
        Classic infantile Pompe patients approaching adulthood: a cohort study on consequences for the brain.
        Dev Med Child Neurol. 2018; 60: 579-586https://doi.org/10.1111/dmcn.13740
        • Korlimarla A.
        • Spiridigliozzi G.A.
        • Crisp K.
        • et al.
        Novel approaches to quantify CNS involvement in children with Pompe disease.
        Neurology. 2020; 95: e718-e732https://doi.org/10.1212/WNL.0000000000009979
        • Ebbink B.J.
        • Poelman E.
        • Plug I.
        • et al.
        Cognitive decline in classic infantile Pompe disease: an underacknowledged challenge.
        Neurology. 2016; 86: 1260-1261https://doi.org/10.1212/WNL.0000000000002523
        • Lee N.C.
        • Peng W.H.
        • Tsai L.K.
        • et al.
        Ultrastructural and diffusion tensor imaging studies reveal axon abnormalities in Pompe disease mice.
        Sci Rep. 2020; 1020239https://doi.org/10.1038/s41598-020-77193-w
        • Gaetani L.
        • Blennow K.
        • Calabresi P.
        • Di Filippo M.
        • Parnetti L.
        • Zetterberg H.
        Neurofilament light chain as a biomarker in neurological disorders.
        J Neurol Neurosurg Psychiatry. 2019; 90: 870-881https://doi.org/10.1136/jnnp-2018-320106
        • Perrot R.
        • Berges R.
        • Bocquet A.
        • Eyer J.
        Review of the multiple aspects of neurofilament functions, and their possible contribution to neurodegeneration.
        Mol Neurobiol. 2008; 38: 27-65https://doi.org/10.1007/s12035-008-8033-0
        • Disanto G.
        • Barro C.
        • Benkert P.
        • et al.
        Serum neurofilament light: a biomarker of neuronal damage in multiple sclerosis.
        Ann Neurol. 2017; 81: 857-870https://doi.org/10.1002/ana.24954
        • Beerepoot S.
        • Heijst H.
        • Roos B.
        • et al.
        Neurofilament light chain and glial fibrillary acidic protein levels in metachromatic leukodystrophy.
        Brain. 2022; 145: 105-118https://doi.org/10.1093/brain/awab304
        • Sánchez-Valle R.
        • Heslegrave A.
        • Foiani M.S.
        • et al.
        Serum neurofilament light levels correlate with severity measures and neurodegeneration markers in autosomal dominant Alzheimer’s disease.
        Alzheimers Res Ther. 2018; 10: 113https://doi.org/10.1186/s13195-018-0439-y
        • Eichler F.
        • Grodd W.
        • Grant E.
        • et al.
        Metachromatic leukodystrophy: a scoring system for brain MR imaging observations.
        AJNR Am J Neuroradiol. 2009; 30: 1893-1897https://doi.org/10.3174/ajnr.A1739
        • Chien Y.H.
        • Tsai W.H.
        • Chang C.L.
        • et al.
        Earlier and higher dosing of alglucosidase alfa improve outcomes in patients with infantile-onset Pompe disease: evidence from real-world experiences.
        Mol Genet Metab Rep. 2020; 23100591https://doi.org/10.1016/j.ymgmr.2020.100591
        • Peng S.S.F.
        • Chien Y.H.
        • Hwu W.L.
        • Liu H.M.
        • Tseng I.W.Y.
        Cerebral diffusion tensor images in infants and neonates with infantile onset Pompe disease.
        Biomed Eng - Appl Basis Commun. 2011; 23: 205-213https://doi.org/10.4015/S1016237211002591
        • Tseng C.H.
        • Chien Y.H.
        • Lee N.C.
        • et al.
        Gene therapy improves brain white matter in aromatic L-amino acid decarboxylase deficiency.
        Ann Neurol. 2019; 85: 644-652https://doi.org/10.1002/ana.25467
        • Kuhle J.
        • Barro C.
        • Andreasson U.
        • et al.
        Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa.
        Clin Chem Lab Med. 2016; 54: 1655-1661https://doi.org/10.1515/cclm-2015-1195
        • Sidman R.L.
        • Taksir T.
        • Fidler J.
        • et al.
        Temporal neuropathologic and behavioral phenotype of 6neo/6neo Pompe disease mice.
        J Neuropathol Exp Neurol. 2008; 67: 803-818https://doi.org/10.1097/NEN.0b013e3181815994
        • Clas P.
        • Groeschel S.
        • Wilke M.
        A semi-automatic algorithm for determining the demyelination load in metachromatic leukodystrophy.
        Acad Radiol. 2012; 19: 26-34https://doi.org/10.1016/j.acra.2011.09.008
        • Tubi M.A.
        • Feingold F.W.
        • Kothapalli D.
        • et al.
        White matter hyperintensities and their relationship to cognition: effects of segmentation algorithm.
        Neuroimage. 2020; 206116327https://doi.org/10.1016/j.neuroimage.2019.116327
        • Broomfield A.
        • Fletcher J.
        • Hensman P.
        • et al.
        Rapidly progressive white matter involvement in early childhood: the expanding phenotype of infantile onset Pompe?.
        JIMD Rep. 2018; 39: 55-62https://doi.org/10.1007/8904_2017_46
        • Rohrbach M.
        • Klein A.
        • Köhli-Wiesner A.
        • et al.
        CRIM-negative infantile Pompe disease: 42-month treatment outcome.
        J Inherit Metab Dis. 2010; 33: 751-757https://doi.org/10.1007/s10545-010-9209-0
        • Zetterberg H.
        • Svenningsson A.
        Serum neurofilament light and prediction of multiple sclerosis in clinically isolated syndrome.
        Neurology. 2019; 92: 313-314https://doi.org/10.1212/WNL.0000000000006906
        • Alirezaei Z.
        • Pourhanifeh M.H.
        • Borran S.
        • Nejati M.
        • Mirzaei H.
        • Hamblin M.R.
        Neurofilament light chain as a biomarker, and correlation with magnetic resonance imaging in diagnosis of CNS-related disorders.
        Mol Neurobiol. 2020; 57: 469-491https://doi.org/10.1007/s12035-019-01698-3
        • Tsai L.K.
        • Hwu W.L.
        • Lee N.C.
        • Huang P.H.
        • Chien Y.H.
        Clinical features of Pompe disease with motor neuronopathy.
        Neuromuscul Disord. 2019; 29: 903-906https://doi.org/10.1016/j.nmd.2019.09.011
        • Filley C.M.
        White matter: organization and functional relevance.
        Neuropsychol Rev. 2010; 20: 158-173https://doi.org/10.1007/s11065-010-9127-9
        • Riley K.J.
        • O’Neill D.P.
        • Kralik S.F.
        Subcortical U-fibers: signposts to the diagnosis of white matter disease.
        Neurographics. 2018; 8: 234-243