AAV9-mediated GAA gene therapy following enzyme replacement therapy discontinuation in children with infantile-onset Pompe disease Authors: Xinting Liu, Yingying Mao, Linyan Hu, Wen He, Xiaodong Wang, Xueyuan Guo, Min Li, Yiru Wang, Yue Wang, Wenhao Ma, Yun Yuan, Zhaoxia Wang, Shuangqing Yu, Shu Zhang, Shiwen Wu, Xiaoyan Dong, Xiaobing Wu, Guang Yang Published: 2026 ABSTRACT SummaryBackgroundWhile our previous research showed that a single infusion of recombinant AAV9-mediated GAA gene therapy enabled infants (<1 year) with infantile-onset Pompe disease (IOPD) and minimal prior enzyme replacement therapy (ERT) to remain ventilator- and ERT-free for at least one year, it was unknown whether the same benefits could be sustained in older children with IOPD with prolonged, high-dose ERT exposure after treatment cessation.MethodsIn this investigator-initiated, single-arm trial (ChiCTR2200065664), six children with IOPD aged 12.5–50.8 months—all having previously received 10–98 cycles of ERT—were enrolled and administered a single intravenous infusion of AAV9-mediated gene therapy. Over a 12-month period, the following outcomes were evaluated: primary efficacy endpoints (gross motor function, and cardiac structure and function); secondary efficacy endpoints (ventilator-free survival, and the proportion of patients who showed improvement at the end of the trial); exploratory endpoints (IOPD-related skeletal muscle pathological changes); safety and tolerability endpoints; and immunogenicity assessment endpoints.FindingsThe intervention was generally well tolerated, with a low incidence of grade ≥3 adverse events (2.31 events per person-time). All six patients remained ventilator-free over the one-year observation period (one patient required transient ventilatory support for pneumonia). However, lung function and cardiac structure showed no further improvement after gene therapy. All children (patient 4 was not documented due to early withdrawal from the study) maintained sustained growth and development throughout follow-up, and overall quality of life showed a trend of improvement (PedsQL 3.0 NMM). During the study period, patient 1 achieved standing with assistance and crawling; patient 2 achieved standing and walking without assistance; and patient 3 achieved walking and sitting without assistance. Muscle biopsies from the three patients revealed relatively high GAA enzyme activity and reduced glycogen accumulation.InterpretationThese preliminary findings suggest that AAV9-mediated GAA gene therapy may reduce reliance on ERT and is associated with improvements in muscle histopathology among biopsied participants.Funding2023YFC3403300; Z231100004823022.Keywords: Infantile-onset Pompe disease (IOPD), Adeno-associated virus serotype 9 (AAV9), Gene therapy, GC301 Research in context Evidence before this study To comprehensively evaluate the current therapeutic landscape, a systematic search was conducted across PubMed, Embase, Cochrane Library, and ClinicalTrials.gov for articles and trial registrations published from database inception to January 1, 2023 using the terms “Pompe disease” AND (“therapy” OR “treatment”) with no language restrictions. ERT with recombinant human acid alpha-glucosidase (rhGAA) has been the standard of care for IOPD since 2006. Despite ERT, most IOPD survivors experience progressive motor, respiratory, and cognitive deficits. Recently, the widespread implementation of newborn screening programs has enabled earlier diagnosis and the timely initiation of treatment, which significantly contributes to better physical and developmental outcomes. Furthermore, the advent of next-generation ERT has shown success in further stabilizing and addressing the motor and respiratory deficits previously observed. However, the requirement for lifelong bi-weekly intravenous infusions, suboptimal tissue biodistribution to the central nervous system (CNS) and skeletal muscle, and treatment burdens still highlight the unmet medical need. Preclinical data in murine and canine models of Pompe disease demonstrated that systemic administration of AAV9 vectors encoding GAA could drive widespread transduction, including in tissues like skeletal and cardiac muscle, leading to sustained enzyme expression, glycogen clearance, and functional improvement. Our preliminary pilot clinical study (ChiCTR2200063229) of recombinant AAV9-GAA (GC301) in treatment-naïve infants with IOPD under one year of age showed promising safety and efficacy, enabling sustained independence from ERT and ventilation over one year. However, no research was performed on the application of systemic AAV9-mediated gene therapy in patients with IOPD older than one year who had already received extended prior ERT—a common real-world clinical scenario. Added value of this study This single-arm, single-center study provides the first clinical evidence that a single intravenous infusion of GC301 is feasible, safe, and can confer therapeutic benefits in children with IOPD (median age 25.5 months) who had received prolonged prior ERT (median 19.5 months). It provides preliminary evidence suggesting that gene therapy can potentially be integrated following established ERT. Key findings include: (1) Sustained ERT- and ventilator-free survival in five of six patients for one year; (2) All documented children exhibited growth and development during the observation period; (3) Patients 1, 2, and 3 achieved key milestones such as unassisted crawling, standing, and walking during the observation period (the independent standing and walking of patient 1 occurred in the 13th month); (4) A consistent improvement in disease-specific quality of life; (5) Exploratory histopathological analyses of percutaneous quadriceps muscle biopsies suggested reductions in lysosomal glycogen accumulation, alongside apparent improvements in muscle integrity and mitochondrial activity; and (6) A manageable safety profile with no adverse events directly attributed to the gene therapy product. Implications of all the available evidence Collectively, the evidence from preclinical studies, the initial trial in treatment-naïve infants, and this ERT-experienced cohort supports the potential of a single-dose, systemic AAV9-GAA gene therapy as a promising therapeutic avenue for IOPD. It suggests a move away from chronic, burdensome enzyme infusions towards a strategy that may mitigate some of the tissue biodistribution limitations inherent to ERT. Introduction Pompe disease is a rare autosomal recessive disorder caused by variants in the GAA gene located on chromosome 17q25.2-q25.3, which encodes the lysosomal hydrolase enzyme, acid alpha-glucosidase (GAA).1, 2, 3 GAA is synthesized as an inactive proenzyme in the endoplasmic reticulum.4, 5, 6 Within the Golgi apparatus, it acquires a mannose 6-phosphate (M6P) lysosomal targeting signal, which facilitates its binding to the mannose 6-phosphate receptor (M6PR) for transport to lysosomes, where it functions to hydrolyze glycogen.7,8 Variants in the GAA gene can disrupt any stage of this finely regulated process—from synthesis and correct folding to lysosomal trafficking and proteolytic activation—resulting in varying degrees of enzyme deficiency that correlate with the clinical severity of the disease.3,9,10 The complete or partial loss of GAA function leads to the pathological accumulation of glycogen within lysosomes, causing lysosomal swelling, rupture, and cascading cellular damage that ultimately results in progressive tissue injury and organ dysfunction.11,12 Traditionally, Pompe disease is classified into two primary phenotypes based on the age of onset, organ involvement, and rate of progression: infantile-onset Pompe disease (IOPD) and late-onset Pompe disease (LOPD).13 IOPD is defined by symptom onset within the first 12 months of life and invariably features hypertrophic cardiomyopathy. Without treatment, this form is typically fatal within the first year or two of life due to cardiorespiratory failure.14,15 LOPD includes patients with onset before 12 months without cardiomyopathy, and all individuals with onset after 12 months.16 It primarily presents with progressive proximal muscle weakness and respiratory insufficiency, while clinically significant cardiac involvement is uncommon.17,18 Enzyme replacement therapy (ERT) with recombinant human GAA (rhGAA) has been the standard of care since its approval in 2006.19 While ERT provides undeniable benefits by prolonging survival and improving motor and respiratory function, it faces limitations. These challenges include: (1) the burden of lifelong, high-frequency intravenous infusions (administered every two weeks); (2) the requirement for high doses of rhGAA (20–40 mg/kg) to achieve therapeutic benefit; (3) prohibitively high costs; and (4) intrinsically low tissue uptake efficacy, particularly in CNS and skeletal muscle.20,21 Consequently, despite an initial favorable response to ERT, disease progression persists, and severe motor, cardiac, respiratory, and cognitive deficits remain prevalent among IOPD survivors.22 To assess the efficacy of an alternative therapy, a recombinant adeno-associated virus serotype 9 (AAV9) vector carrying a codon-optimized human GAA gene expression cassette, designated GC301, was engineered to enhance tissue transduction efficiency.2 Preliminary pilot clinical studies have indicated that a single intravenous injection of GC301 enabled patients with IOPD under one year of age to remain free from ERT and ventilator support for at least a one-year observation period, while demonstrating a favorable safety profile. In the present study, we aimed to determine the safety and efficacy of GC301 in patients with IOPD who were older than one year and had received prolonged ERT. Methods Study design and ethical issues This investigator-initiated trial (ChiCTR2200065664) evaluated the safety and efficacy of GC301 treatment in patients with IOPD aged 1–6 years who had received prior prolonged ERT (clinically defined in this study as a sufficient duration of prior ERT to establish a stable baseline without rapid disease progression). The protocol was approved by the Chinese PLA General Hospital Ethics Committee (S2022-596-04) and adhered to the principles of the Declaration of Helsinki, the International Council for Harmonisation Good Clinical Practice (ICH-GCP) guideline, and the Good Clinical Practice guideline of the National Medical Products Administration of China (NMPA-GCP). Prior to enrollment, written informed consent was obtained from all legal guardians. For an exploratory, investigator-initiated trial with a limited sample size and an open-label, single-arm design, a rigorous internal monitoring process was implemented to ensure participant safety and data integrity. This process included the following elements. 1.Direct monitoring by the principal investigator and research team: The primary responsibility for safety oversight rested with the research team led by the Principal Investigator. The team conducted close clinical follow-up for all participants and reviewed all safety data in real time. Direct monitoring by the principal investigator and research team: The primary responsibility for safety oversight rested with the research team led by the Principal Investigator. The team conducted close clinical follow-up for all participants and reviewed all safety data in real time. 2.Predefined protocol stopping rules: The study protocol included clearly defined criteria for safety pauses and study termination (see Supplementary Material, “Trial Termination Criteria” section). Predefined protocol stopping rules: The study protocol included clearly defined criteria for safety pauses and study termination (see Supplementary Material, “Trial Termination Criteria” section). 3.Institutional internal review mechanisms: The study underwent periodic continuing review by the Institutional Review Board (the Ethics Committee of the Chinese PLA General Hospital). In addition, the research team held regular internal safety review meetings to systematically evaluate cumulative safety information. Institutional internal review mechanisms: The study underwent periodic continuing review by the Institutional Review Board (the Ethics Committee of the Chinese PLA General Hospital). In addition, the research team held regular internal safety review meetings to systematically evaluate cumulative safety information. Participants Patients with confirmed IOPD were enrolled between April and October 2023. The key inclusion criteria were: (1) Aged 1–6 years, regardless of sex; (2) Onset of Pompe disease symptoms before 12 months of age, with IOPD diagnosis confirmed by GAA gene mutation analysis, deficient GAA enzyme activity, and classic clinical features of IOPD (specifically, a documented history of hypertrophic cardiomyopathy and muscle weakness/hypotonia prior to 12 months of age)23; (3) Prior receipt of ERT with clinical stability, defined as the absence of rapid disease progression (e.g., no acute decline in cardiopulmonary function, such as a new requirement for respiratory support, and no rapid loss of motor milestones prior to screening); (4) Ability to complete motor function assessments; (5) Legal guardian(s) capable of understanding and adhering to the study protocol, and who voluntarily provided signed informed consent for the participant's enrollment. The exclusion criteria were: (1) Clinically significant organic diseases other than those attributable to IOPD, including significant cardiovascular, pulmonary, hepatic, renal, or neurological disorders, or other severe comorbidities; (2) Requirement of invasive respiratory support or non-invasive respiratory support for >16 h per day; (3) Severe non-respiratory illness requiring systemic therapy and/or hospitalization within 2 weeks before screening; (4) Upper or lower respiratory tract infection requiring antiviral/antibacterial therapy and/or hospitalization within 4 weeks before screening; (5) Use of immunomodulatory therapy (excluding inhaled corticosteroids) within 90 days prior to screening; (6) History of severe adverse reaction to corticosteroids; (7) Serum anti-AAV9 neutralizing antibody titer >1:100; (8) Previous or concurrent participation in another gene therapy clinical trial; (9) Any other condition deemed by the investigator to render the participant unsuitable for the trial. Intervention, outcomes and assessments Patients received a single intravenous infusion of GC301 (driven by a CMV enhancer/chicken β-actin promoter) at a dose of 1.2 × 1014 vg/kg following a 4–20 day washout period from ERT.2 Oral prednisolone (2 mg/kg/day starting Day −1) was tapered over 8 weeks to mitigate immune responses. The safety was assessed by the incidence of adverse events (graded per Common Terminology Criteria for Adverse Events Version 5.0) and serious adverse events. Assessments of efficacy (defined as ventilator-free survival and independence from ERT at 52 weeks). Motor function will be evaluated using a multi-faceted approach, including the Gross Motor Function Measure-88 (GMFM-88) total score (%), the attainment of World Health Organization (WHO) motor milestones, and imaging studies such as muscle ultrasonography. Cardiac function will be monitored via echocardiography, with a primary focus on the change in the height-adjusted left ventricular mass Z-score and left ventricular ejection fraction over the 52-week period. Pulmonary function will be assessed using tPTEF/tE (%) and VT/kg (ml/kg). The health-related quality of life will be compared before and after treatment using the PedsQL 3.0 Neuromuscular Module (PedsQL 3.0 NMM). Growth metrics (height and weight) will be evaluated against age- and sex-specific normative growth charts for Chinese children and adolescents aged 0–18 years. The circulating biomarkers to be monitored include creatine kinase (CK), alanine aminotransferase (ALT), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH), with measurements performed by the hospital's clinical laboratory. Other major monitoring items are elaborated in detail in this method section, focusing on the efficacy, mechanisms, safety, and durability of the gene therapy. Details of the regimen are shown in Fig. 1. Clinical trial design. A single-arm, single-center exploratory interventional study with a one-year post-treatment observation period. A. A defined process for patient enrollment and exclusion. IQR, interquartile range. B. A gene therapy strategy for efficacy and safety monitoring. The dosing regimen begins with a cessation of ERT, after which enrolled patients, at baseline (Week 0), receive oral prednisolone prophylaxis at 2 mg/kg/day followed by a single intravenous infusion of GC301 at a dose of 1.2 × 1014 vg/kg. Prednisolone is maintained orally at 2 mg/kg/day on the first two days post-infusion, then reduced to 1 mg/kg/day from Day 3 to Day 30. Beginning on Day 31, the prednisolone dose is tapered over 29 days based on clinical assessment, a strategy aimed at mitigating immune responses to the AAV vector and maximizing transduction efficiency. Comprehensive safety and efficacy data are systematically collected at scheduled timepoints, including biochemical profiles, cardiac function, respiratory capacity, motor function, pharmacokinetics, and immunological responses to the therapy. Recombinant AAV9 (rAAV9-coGAA) description GC301 (rAAV9-coGAA) is a recombinant AAV9 vector carrying a codon-optimized human GAA gene and is produced using the baculovirus-insect cell system.24 The expression of the therapeutic transgene is driven by a hybrid promoter consisting of the cytomegalovirus (CMV) early enhancer element and the chicken beta-actin (CAG) promoter. Droplet digital PCR (ddPCR) assay for vector genome in blood The vector genome copy number in whole blood samples was quantified using a ddPCR assay. Blood samples were collected in EDTA-K2 tubes, and genomic DNA was extracted from 100 μL of whole blood using a magnetic Universal Genomic DNA Kit (TIANGEN, China), according to the manufacturer's protocol. The ddPCR assay employed the same coGAA-specific primers and probe used for the quantification of the rAAV9-coGAA vector. The sequences were as follows: forward primer 5′-GGCACCTGGTCGTGGAACTC-3′, reverse primer 5′-CCAGCCTCTGATCGGCAAGG-3′, and a FAM-labeled probe 5′-TGGCCAGGCTCCACCGCCTT-3′ with a TAMRA quencher (synthesized by Sangon Biotech, Shanghai, China). Each sample was analyzed in duplicate. The results, expressed as the genome copy number per milliliter of blood, demonstrated a lower limit of detection for the assay of 1.76 × 103 copies/mL. Detection of cross-reactive immunological material (CRIM) Patient peripheral blood mononuclear cells (PBMCs) were collected approximately 10 days after the most recent ERT session and prior to the GC301 infusion. These PBMCs were then immortalized by Epstein–Barr virus (EBV) transformation to generate lymphoblastoid cell lines (LCLs). The LCLs were cultured and passaged for a minimum of 10 passages before analysis. For Western blot analysis, LCLs were lysed using RIPA buffer, and 40 μg of total protein was loaded per lane. Protein detection was carried out using a mouse monoclonal antibody against human GAA capable of recognizing both the precursor and active forms of the protein. β-Actin served as the loading control, and Raji cell protein lysate was included as a positive control. Samples were classified as cross-reactive immunologic material (CRIM)-positive upon detection of any GAA precursor or mature protein fragments.25 Detection of anti-AAV9 nAbs A strictly validated cell-based microneutralization (MN) assay was employed to quantify anti-AAV9 neutralizing antibodies at the bioanalytical laboratory of Beijing Genecradle Therapeutics Inc. Serum or plasma samples were pre-treated at 56 °C for 30 min to inactivate complement. Starting from a 1:20 dilution, 50 μL of each sample was subjected to a series of two-fold dilutions. These dilutions were then incubated with 2 × 108 vector genomes (vg) of rAAV9-EGFP-2A-Gluc in 50 μL of DMEM supplemented with 0.1% BSA for 1 h at 37 °C. The virus-serum mixture was subsequently transferred onto HEK-293-C340 cells (a highly permissive subclone) seeded in 96-well plates containing D10 medium with 1 mM sodium butyrate, which enhances transduction. The plates were incubated at 37 °C under 5% CO2 for 48–72 h. After incubation, Gaussian luciferase activity was measured in relative light units (RLU). Wells with a coefficient of variation (CV%) for RLU exceeding 30% were excluded from the analysis. The percentage of transduction inhibition (TI) was calculated using the formula: TI% = [1 − (mean RLUtest − mean RLUcc)/(mean RLUvc − mean RLUcc)] × 100%, where “test” is the test sample, “vc” is the virus control, and “cc” is the cell control. The half-maximal inhibitory concentration (IC50) titer, representing the serum dilution that causes a 50% reduction in transduction, was determined through curve-fit modeling. Detection of anti-AAV9 binding antibodies An antigen-binding enzyme-linked immunosorbent assay (ELISA) was used to detect antibodies that bind to the AAV9 capsid. 96-well plates were coated with 50 ng per well of AAV9 vector particles in carbonate buffer (pH 9.6) and incubated overnight at 4 °C. After coating, the plates were blocked with PBS containing 5% skim milk for 2 h at 37 °C. For system suitability control, low and high-quality positive controls were prepared using a mouse monoclonal anti-AAV9 antibody at concentrations of 0.12 μg/mL and 5 μg/mL, respectively, diluted in pooled negative serum. Test samples and controls were diluted to a minimum required dilution (MRD) of 1:100 in 1% skim milk. For the confirmatory assay, samples were pre-incubated with 50 ng of soluble AAV9 vector particles for 1 h to assess specificity through competitive inhibition. The samples were then added to the coated plates in duplicate and incubated at 37 °C for 1 h. Following washes, a mixture of horseradish peroxidase (HRP)-conjugated goat anti-human and goat anti-mouse IgG antibodies was added. The signal was developed using tetramethylbenzidine substrate, and the optical density (OD) was read at 450 nm. The signal-to-noise (S/N) ratio was calculated by dividing the OD value of the sample by the OD value of the negative control. The percentage inhibition in the confirmatory assay was defined as [1 − (S/N ratio with AAV9 particles/S/N ratio without AAV9 particles)] × 100%. A sample was considered positive if its inhibition percentage exceeded the confirmatory cut point of 43.84%. The binding antibody titer was reported as the reciprocal of the maximum serum dilution yielding a positive result. Detection of anti-GAA binding antibodies The assay for detecting antibodies against the human GAA followed a similar ELISA protocol to the anti-AAV9 binding antibody assay, with the following modifications: (1) the coating antigen was full-length human GAA protein expressed in CHO cells; (2) the positive control was a rabbit monoclonal anti-GAA antibody used at 0.1 μg/mL and 3 μg/mL; (3) the detection antibody was a mixture of HRP-conjugated goat anti-human and goat anti-rabbit IgG. The confirmatory cut point for a positive result was set at an inhibition percentage of ≥38.512%. The assay's lower limit of detection (LLOD) was 29.67 ng/mL. Detection of cellular immune responses to AAV9 capsid and GAA Cellular immune responses were evaluated using an enzyme-linked immunospot (ELISpot) assay to detect interferon-gamma (IFN-γ) secretion upon stimulation with peptide pools. Overlapping peptides (18–20 amino acids in length, with 10–12 amino acid overlaps) covering the full sequence of the AAV9 VP1 capsid protein and human acid alpha-glucosidase (GAA) were synthesized by ChinaPeptide Qyaobio (Shanghai, China) and Scilight-Peptide (Beijing, China), respectively. Fresh PBMCs were isolated using BD Vacutainer® CPT™ mononuclear cell preparation tubes according to the manufacturer's instructions. The cells were washed and resuspended at a concentration of 2 × 106 cells/mL in culture medium. A total of 100 μL of the cell suspension was added to each well of a pre-coated 96-well plate and stimulated with 2.5 μg/mL of the respective peptide pools. Unstimulated cells (in medium alone) served as the negative control, and cells stimulated with an anti-CD3 monoclonal antibody were included as the positive control. After a 36-h incubation period at 37 °C under 5% CO2, the cells were removed, and the plate was processed according to the kit protocol (Human IFN-γ [ALP] ELISpot kit; Mabtech, Sweden). This included incubation with a biotinylated detection antibody, followed by an enzyme conjugate, and development with BCIP/NBT substrate. Spot-forming units (SFUs) were counted manually. A response was considered positive if the number of SFUs per million cells exceeded 50 and was at least three times greater than the value observed in the unstimulated negative control wells. Muscle imaging Muscle ultrasound was performed on the quadriceps femoris, and echogenicity was graded using a modified Heckmatt scale,26 which classifies findings into the following three levels: Grade 1 (Normal): Muscle echogenicity is not increased relative to the adjacent epimysium, and the perimysium remains sharply defined. Grade 2 (Mild Increase): Less than 50% increase in muscle echogenicity compared to the superficial epimysium, which remains clearly visible. Grade 3 (Marked Increase): Greater than 50% increase in muscle echogenicity compared to the superficial epimysium, accompanied by a loss of epimysial clarity. Histochemical analysis Percutaneous muscle biopsies were performed on patients 3, 5, and 6 at baseline and 26 weeks post-treatment. Following local anesthesia with 2% lidocaine (without direct muscle infiltration to avoid altering tissue biochemistry), a 1-cm longitudinal incision was made. A cylindrical muscle sample was excised, immediately placed on saline-moistened gauze, and maintained on ice to preserve enzymatic reactivity and prevent macromolecule degradation. Periodic Acid-Schiff (PAS) Staining: Paraffin sections (4 μm) were fixed in ice-cold 4% paraformaldehyde for 10 min and rinsed three times in PBS. Slides were oxidized with 0.5% periodic acid for 10 min at room temperature, thoroughly washed in distilled water (three changes, 2 min each), and incubated in Schiff's reagent for 15–20 min in the dark. After a 5-min rinse under running water, nuclei were counterstained with modified Harris's hematoxylin for 2 min, differentiated in 1% acid alcohol for 30 s, and blued in running water for 5 min. Sections were dehydrated through a graded ethanol series (70%–100%), cleared in xylene, and mounted with neutral balsam. Glycogen content, visualized by PAS staining, was compared before and after GC301 injection using transmission electron microscopy for ultrastructural assessment. Succinate Dehydrogenase (SDH) Staining: Sections were air-dried for 30 min at room temperature and then incubated at 37 °C for 45 min in a freshly prepared solution containing 0.2 M phosphate buffer (pH 7.6), 0.05 M sodium succinate, and 1 mM nitroblue tetrazolium, protected from light. After three 1-min rinses in 0.1 M phosphate buffer, slides were dehydrated in graded acetone (30%, 60%, 90%; 2 min each), cleared in xylene, and mounted with synthetic resin. Mitochondria-rich, oxidative Type I fibers stained dark blue-purple and were imaged at × 20 magnification for fiber-type quantification. Oil Red O Staining: Following 10 min of fixation in 4% paraformaldehyde and three PBS washes, 4-μm cryosections were immersed in 60% isopropanol for 2 min and stained for 15 min in the dark with freshly prepared modified Oil Red O solution (stock solution to distilled water ratio of 3:2). Differentiation in 60% isopropanol (3–5 s) was stopped when lipid droplets appeared bright red against a colorless stroma. Sections were rinsed in running water for 5 min, counterstained with modified Harris's hematoxylin for 1 min, differentiated in 1% acid alcohol for 30 s, and blued under tap water. After dehydration through graded ethanol, clearing in xylene, and mounting in glycerol gelatin, slides were imaged at × 20 magnification. Lipid droplets were quantified as a percentage area using ImageJ software. Desmin Immunostaining: Cryosections were fixed with 4% paraformaldehyde for 15 min and washed three times (5 min each) in PBS to remove excess fixative. Endogenous peroxidase activity was quenched by incubation in 3% H2O2 in methanol containing 0.3% NaN3 for 15 min at room temperature. After three additional PBS rinses, sections were blocked for 1 h in 2% normal goat serum diluted in PBS. They were then incubated overnight at 4 °C with a mouse anti-desmin monoclonal antibody (1:5,000, clone MA5-13259, Thermo Fisher) prepared in blocking buffer. Following three PBS washes, sections were incubated with Alexa Fluor 488-conjugated goat anti-mouse IgG (1:500, Invitrogen) for 1 h at room temperature. Nuclei were counterstained with DAPI for 5 min, washed in PBS, and slides were mounted with ProLong Gold antifade medium. Images were acquired on a confocal microscope at ×20 magnification. GAA enzyme activity assay GAA Enzyme Activity Assay in Dried Blood Spots (DBS).27 A 3.2 mm punch from a DBS was placed into a 96-well plate. An assay cocktail containing the specific substrate and internal standard was added to each well. The plate was sealed with an aluminum film and incubated for 18 h at 37 °C with orbital shaking. Following incubation, the reaction was quenched by adding a quench solvent, and the solution was thoroughly mixed before being transferred to a deep-well plate. A liquid–liquid extraction by adding ethyl acetate and water to each well. After centrifugation, the upper organic layer (ethyl acetate phase) was carefully transferred to a new plate. The solvent was evaporated under a stream of nitrogen gas, and the residue was reconstituted in a mobile phase solution prior to analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The analysis was conducted using a Waters Acquity I-Class UPLC system (Waters, MA) coupled to a XevoQ-S mass spectrometer (Waters, MA), equipped with a C18 analytical column (ThermoFisher, Shanghai, China). GAA Enzyme Activity Assay in Muscle Tissue.12 Muscle tissue samples, collected from the quadriceps, were immediately flash-frozen in liquid nitrogen and stored at −70 °C until analysis. The frozen tissues were homogenized on ice using stainless-steel beads in a RIPA lysis buffer supplemented with protease inhibitors. The total protein concentration of the homogenate was quantified using the Bradford assay. To measure GAA activity, the tissue homogenates were incubated with a reaction mix containing a specific substrate (e.g., 4-MU-α-glucopyranoside) and an internal standard at 37 °C for 2 h. The reaction was terminated by adding methanol, followed by vigorous vortex mixing. The mixture was then centrifuged at 13,300 rpm at room temperature. The resulting supernatant was collected and transferred vials containing methanol for analysis. Chromatographic separation was achieved using an Agilent 1290 UPLC system (Agilent Technologies, Beijing, China) coupled to a 6475 triple-quadrupole mass spectrometer (Agilent Technologies, Beijing, China). A C18 column was used, and the mass spectrometer was operated in positive electrospray ionization (ESI) mode with multiple reaction monitoring (MRM) for detection. Western blot analysis of GAA protein expression in muscle tissue Muscle tissue samples were homogenized on ice in RIPA lysis buffer supplemented with protease inhibitors using stainless steel beads. The total protein concentration of the resulting lysate was determined using the Bradford assay, with bovine serum albumin serving as the standard. Equal amounts of protein (30 μg per lane) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 12.5% gels. Following electrophoresis, the proteins were electrophoretically transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were then blocked with 5% (w/v) non-fat dry milk prepared in PBS containing 0.1% Tween-20 (PBST) for 1 h at room temperature to prevent non-specific antibody binding. For immunodetection, the membranes were incubated overnight at 2–8 °C with a primary mouse monoclonal antibody against human GAA (clone 4B1, generated in-house) at a concentration of 2 μg/mL in blocking buffer. After three rigorous 10-min washes with PBST (>30 mL per wash), the membranes were incubated for 1 h at room temperature with 0.16 μg/mL horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG secondary antibody (ZSGB-BIO, #ZB-2305, Beijing, China) diluted in 5% milk-PBST. Following another series of washes, the antigen–antibody complexes were visualized using an enhanced chemiluminescence (ECL) substrate. The membranes were exposed for 180 s, and the resulting chemiluminescent signals were captured. To ensure equal protein loading across lanes, the membrane was subsequently probed for GAPDH, which served as the housekeeping Densitometric analysis of the protein bands was performed using ImageJ software (National Institutes of Health, MA) for quantification. Adverse event and serious adverse event In clinical research, adverse events (AEs) are defined as any unfavorable medical occurrence in a patient or clinical trial subject administered a medicinal product, which does not necessarily have a causal relationship with the treatment. The severity of AEs is systematically classified according to the National Cancer Institute's Common Terminology Criteria for Adverse Events (CTCAE), version 5.0 and above,28 into five grades: Grade 1 (Mild) includes asymptomatic or mild symptoms requiring only clinical observation and no treatment, with specific laboratory thresholds such as AST/ALT levels between the upper limit of normal (ULN) and 3 times the ULN; Grade 2 (Moderate) involves symptoms requiring minimal, local, or non-invasive intervention, and limits instrumental activities of daily living (ADL), with AST/ALT levels between 3 and 5 times the ULN; Grade 3 (Severe) includes medically significant events that require hospitalization or prolongation of existing hospitalization, are disabling, or limit self-care ADL, with AST/ALT levels between 5 and 20 times the ULN; Grade 4 (Life-threatening) entails urgent intervention for life-threatening consequences, with AST/ALT levels exceeding 20 times the ULN; and Grade 5 (Death) is fatal and directly related to the AE. Separately, a Serious Adverse Event (SAE) is any untoward medical occurrence that, at any dose, results in death, is life-threatening, requires inpatient hospitalization or prolongs existing hospitalization, results in persistent or significant disability/incapacity, constitutes a congenital anomaly/birth defect, or may jeopardize the subject and require medical or surgical intervention to prevent one of these outcomes.29 Statistical strategy Pre-specified analyses: As per our study protocol (see Supplementary Material, “Statistical Analysis” section), descriptive statistical analyses were conducted for the primary efficacy endpoints (GMFM-88 scores, LVM Z-scores, and LVEF), safety, efficacy, pharmacokinetic, and immunogenicity data related to GC301. Exploratory (post-hoc) analyses: Given the longitudinal, multi-timepoint data generated over the one-year follow-up, several post-hoc analyses were performed for further data exploration. These included: (1) Linear regression analyses to estimate individual longitudinal trends for all clinical continuous variables; (2) Trend analyses of concentration changes in clinical continuous variables over the follow-up period, with the slope and its 95% confidence interval (95% CI) reported as an indicator of potential treatment effect. In addition, the analysis of Periodic Acid-Schiff (PAS) staining in muscle biopsy tissues—using the Mann–Whitney U test with a two-sided significance level of p < 0.05 to compare pre- and post-treatment timepoints—was also performed as a post-hoc analysis. All statistical analyses were performed using GraphPad Prism, version 10 (GraphPad Software, San Diego, CA, USA). For all analyses, a nominal two-sided alpha level of 0.05 was applied. Role of the funding source Funding for this study was provided by the National Key R&D Program of China (Grant No. 2023YFC3403300) and the Beijing Science and Technology Plan Project (Grant No. Z231100004823022). These grants covered the costs associated with vector manufacturing and quality control for the clinical trials, which were conducted by Beijing Genecradle Therapeutics Inc., the developer of GC301. The funders had no role in the study design; data collection, analysis, or interpretation; manuscript writing; or the decision to submit the manuscript for publication. All academic investigators retained full independent control over data access, data analysis and interpretation, and the decision to publish. Results Patient enrollment A total of 14 patients with IOPD met the initial enrollment criteria. Of these, eight patients were excluded for the following reasons: four due to anti-AAV9 neutralizing antibody titers and/or anti-GAA antibody titers exceeding the threshold (>1:100); and the investigator assessed another four patients as unsuitable for participation in this clinical trial (poor overall functional status (n = 1), and motor function development comparable to that of age-matched typically developing children (n = 3)). Consequently, six patients were ultimately enrolled in gene therapy. However, patient 4 was withdrawn from the study after developing severe pneumonia in the ninth week post-enrollment, which necessitated resumption of ERT in accordance with clinical guidelines. Therefore, only five patients completed the full 52-week follow-up period. For this report, data from patient 4 collected over the first nine weeks were included in the analysis, including the adverse events statistics (Fig. 1A). This gene therapy strategy outlines a structured approach for monitoring safety and efficacy (Fig. 1B). The dosing regimen initiates with an ERT washout period. At baseline (Week 0), enrolled patients receive oral prednisolone prophylaxis (2 mg/kg/day) followed by a single intravenous infusion of GC301 at 1.2 × 1014 vg/kg. Post-infusion, the prednisolone regimen is designed to mitigate immune responses against the AAV vector and to maximize transduction efficiency (Fig. 1B). Baseline characteristics Table 1 summarizes the baseline clinical characteristics of the six enrolled patients. The cohort consisted of three boys and three girls, with a median age of 25.5 (range: 12.5–50.8) months. All patients showed signs of growth and developmental delay, as indicated by their median height and weight of 85.5 (range: 74.5–102.0) cm and 10.0 (range: 7.8–13.1) kg, respectively. The median age at symptom onset was 4.0 (range: 1.0–10.0) months. The median age at the initiation of ERT was 4.9 (range: 3.3–11.7) months, with patients receiving a median of 30 ERT infusions over a median duration of 19.5 (range: 3.8–48.5) months prior to enrollment. Table 1Baseline clinical characteristics of enrolled patients.Patient 1Patient 2Patient 3Patient 4Patient 5Patient 6GenderMaleMaleFemaleFemaleMaleFemaleAge at enrollment (Month)12.516.827.050.250.824.0GAA variantc.2662G > T (p.E888∗)c.2662G > T (p.E888∗)c.1411-1414delGAGA (p.E471Pfs∗5)c.859-2A > Tc.2662G > T (p.E888∗)c.1562A > T (p.E521V)c.1942G > A (p.G648S)c.1082C > T (p.P361L)c.1802C > G (p.S601W)c.1861T > G (p.W621G)c.1935C > A (p.D645E)c.2662G > T (p.E888∗)Height (cm)74.578.084.0102.0100.087.0Weight (kg)7.88.89.812.213.110.2Age at onset (Month)3107151Age at initiation of ERT (Month)3.911.78.03.35.54.3Total number of ERT before gene therapy201019988540Dosage and frequency of ERT20 mg/kg, q2w20 mg/kg, q2w15 mg/kg, qm32 mg/kg, q2w20 mg/kg, q2w20 mg/kg, q2wERT duration (Month)8.85.018.948.541.520.0Days from the last ERT to gene therapy814206124Blood GAA activity (μmol/liter/hr)82.82.11.51.73.22.5Age at gene therapy (Month)12.917.127.652.051.024.7LVM Z-scores3.72.63.9−0.5−2.12.4LVEF (%)766773677865Modified Heckmatt ultrasound score11128111111GMFM-88 (%)35.431.554.458.821.115.8VT/kg (ml/kg)8.57.27.19.97.56.7tPTEF/tE (%)14.99.112.223.224.926.8PedsQL™ 3.0 NMM scores70.586.060.055.045.045.0ALT (U/L)161.6146.9194.5120.996.8211.1AST (U/L)456.1176.1407.5192158.9265.5LDH (U/L)1165.9604.81633451.6758.7972.1CRIM status++++++AAV-binding antibodies––––––Anti-AAV9 nAb––––––Anti-GAA binding antibodies––––––ALT, Alanine aminotransferase (normal range: 0−40 U/L); AST, aspartate aminotransferase (normal range: 0−40 U/L); CRIM status, cross-reactive immunologic material; GMFM-88, Gross Motor Function Measure-88; LDH, lactate dehydrogenase (normal range: 40−250 U/L); LVEF, left ventricular ejection fraction (normal range: >50%); LVM Z-scores, left ventricular mass Z-scores (normal range: −2 to +2); tPTEF/tE, the time to reach peak tidal expiratory flow as a proportion of total expiratory time (normal range: severe obstructive ventilation <15% pattern; moderate 15–22%; and mild 23–27%); VT/kg, weight-based tidal volume (normal range: >6 mL/kg); +, positive; -, negative; q2w, every 2 weeks; qm, every month. Baseline clinical characteristics of enrolled patients. ALT, Alanine aminotransferase (normal range: 0−40 U/L); AST, aspartate aminotransferase (normal range: 0−40 U/L); CRIM status, cross-reactive immunologic material; GMFM-88, Gross Motor Function Measure-88; LDH, lactate dehydrogenase (normal range: 40−250 U/L); LVEF, left ventricular ejection fraction (normal range: >50%); LVM Z-scores, left ventricular mass Z-scores (normal range: −2 to +2); tPTEF/tE, the time to reach peak tidal expiratory flow as a proportion of total expiratory time (normal range: severe obstructive ventilation <15% pattern; moderate 15–22%; and mild 23–27%); VT/kg, weight-based tidal volume (normal range: >6 mL/kg); +, positive; -, negative; q2w, every 2 weeks; qm, every month. The clinical presentation of these patients was consistent with IOPD. Cardiac assessment revealed myocardial hypertrophy in patients 1, 2, 3, and 6, while patients 4 and 5 did not show myocardial hypertrophy at baseline. Signs of restrictive ventilatory dysfunction were indicated by reduced ventilation efficiency, with a median tidal volume per kilogram (VT/kg) of 7.4 (range: 6.7–9.9) mL/kg for the cohort. Furthermore, patients 1, 2, and 3 demonstrated clear respiratory muscle weakness, shown by their low ratio of time to peak tidal expiratory flow to total expiratory time (tpTEF/tE), which ranged from 9.1% to 14.9%. Motor function and quality of life were substantially impacted. Assessments using the Gross Motor Function Measure-88 (GMFM-88) revealed substantial impairments in gross motor skills, with scores ranging from 15.8% to 58.8%. Correspondingly, the Pediatric Quality of Life Inventory™ Neuromuscular Module (PedsQL™ 3.0 NMM) scores were between 45.0 and 86.0, reflecting a moderate to severe impact on quality of life. Laboratory findings revealed elevated levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH), which are consistent with persistent muscle and liver injury. The cohort exhibited a spectrum of pathogenic variants in the GAA gene. The most recurrent variant observed was c.2662G > T (p.E888∗), present in four patients (patients 1, 2, 5, and 6). Other identified variants included c.1411-1414delGAGA (p. E471Pfs∗5), c.859-2A > T, c.1562A > T (p.E521V), c.1942G > A (p.G648S), c.1082C > T (p.P361L), c.1802C > G (p.S601W), and c.1935C > A (p.D645E). All patients were cross-reactive immunologic material (CRIM)-positive and tested negative for anti-AAV9 neutralizing antibodies and anti-GAA antibodies at baseline (Table 1). Safety of GC 301 therapy The safety profile of GC301 gene therapy was assessed over a one-year follow-up period (Table 2). A total of 105 adverse events were recorded across 5.2 person-year of observation. The analysis of these events revealed that 2.31 events per person-time were classified as grade >3 in severity, 5.58 events per person-time were assessed as possibly related to the drug therapy, and 2.69 events per person-time were serious adverse events. The most commonly recorded adverse events were upper respiratory infections and/or transient dyslipidemia. Table 2Adverse events during the 52-week observation period following GC301 treatment.Number of patientsNumber of eventsNumber of events (grade ≥3)Possibly drug-related adverse eventsSerious adverse eventsFever67020Upper respiratory tract infection522502Pneumonia and bronchitis514409Influenza33103Cough11000ALT increased11010AST increased56050ALP increased11100Hypertriglyceridemia48080Hypercholesterolemia5130130Constipation11000Diarrhea25100Functional gastrointestinal disorder44000Eczema22000Aphtha25000Anaemia12000Hypoxemia22000Hypercapnia12000Acidosis24000Urinary tract infection11000Right leg movement restricted11000Rate55 (10.58)105 (20.19)12 (2.31)29 (5.58)14 (2.69)According to the Common Terminology Criteria for Adverse Events 5.0, Grade 1 Mild; asymptomatic or mild symptoms; clinical or diagnostic observations only; intervention not indicated. Grade 2 Moderate; minimal, local or noninvasive intervention indicated; limiting age-appropriate instrumental activities of daily living. Grade 3 Severe or medically significant but not immediately life-threatening; hospitalization or prolongation of hospitalization indicated; disabling; limiting self-care activities of daily living. Grade 4 Life-threatening consequences; urgent intervention indicated. Grade 5 Death related to an adverse event. Incidence rate of adverse events was calculated as the number of events divided by the total person-time (5.2 person-year: five patients for 1 year, and one patient for 0.2 years) of follow-up. Adverse events during the 52-week observation period following GC301 treatment. According to the Common Terminology Criteria for Adverse Events 5.0, Grade 1 Mild; asymptomatic or mild symptoms; clinical or diagnostic observations only; intervention not indicated. Grade 2 Moderate; minimal, local or noninvasive intervention indicated; limiting age-appropriate instrumental activities of daily living. Grade 3 Severe or medically significant but not immediately life-threatening; hospitalization or prolongation of hospitalization indicated; disabling; limiting self-care activities of daily living. Grade 4 Life-threatening consequences; urgent intervention indicated. Grade 5 Death related to an adverse event. Incidence rate of adverse events was calculated as the number of events divided by the total person-time (5.2 person-year: five patients for 1 year, and one patient for 0.2 years) of follow-up. Survival All six patients survived the 52-week follow-up period. Prior to study enrollment, all six patients were completely independent of ventilatory support; five of the six patients remained free of ventilator support and maintained complete independence from ERT following GC301 administration. One patient (patient 4) was withdrawn from the study on day 60 post-treatment due to a severe episode of pneumonia. This event necessitated mechanical ventilation and the resumption of ERT as part of standard clinical care. However, following recovery, the patient was successfully weaned from ventilatory support and discontinued ERT a second time after seven months of re-initiation, ultimately achieving sustained respiratory independence for the remainder of the 52-week period. Throughout the 52-week observation period, the patients' blood pH, partial pressure of arterial oxygen (PaO2), partial pressure of arterial carbon dioxide (PaCO2), and arterial oxygen saturation (SaO2) generally remained at stable and clinically acceptable levels (Supplementary Table S1). Changes in cardiac structure and function All patients maintained a LVEF above 50% throughout the entire observation period. Following GC301 gene therapy, the median of the patient group showed a slight declining trend in LVEF, however, the slope (−0.013) is too slight to be of clinical significance (Fig. 2A). Regarding left ventricular mass (LVM) Z-scores, patients 1, 3, and 5, as well as the group median, displayed a slight upward trend (slopes: 0.008, 0.005, 0.011, and 0.004, respectively). Similarly, this variation is too slight to be of clinical significance. Although patient 2 showed a slight declining trend (slope: −0.005), his left ventricular mass Z-score has consistently remained within the normal range (Fig. 2B). Furthermore, 12-lead electrocardiograms (ECGs) throughout the entire observation period consistently demonstrated typical features of Pompe disease, such as shortened PR intervals, left ventricular high voltage, and secondary ST-T changes. Overall cardiac parameters remained stable, with no occurrences of malignant arrhythmias or acute cardiac decompensation events. Longitudinal outcomes following GC301 administration. This composite figure presents longitudinal data from six pediatric patients with IOPD treated with GC301. Individual patient trajectories are represented by distinct colored lines (patients 1–6), with cohort medians indicated by black lines. A. LVEF over time. B. LVM Z-score over time. C. Changes in ventilation efficiency (VT/kg) over one year. Tidal volume per kilogram of body weight, reflecting the effective ventilation volume per breath. D. Alterations in respiratory muscle weakness index (tpTEF/tE%): a decrease in this ratio indicates weakness of the respiratory muscles (primarily the diaphragm) during exhalation, leading to limited expiratory flow. E to H. Weight-for-age (E and F) and height-for-age (G and H) growth charts. Individual patient growth trajectories for boys (E and G) and girls (F and H) are plotted against WHO standard percentiles (dashed lines). I. Assesses gross motor abilities in one year follow-up using Gross Motor Function Measure (GMFM-88%). J. Motor milestones (such as sitting, standing, walking) recording, the age in months when the corresponding milestone appears is indicated in parentheses. K. Quality of Life-Neuromuscular Module (PedsQL™ 3.0 NMM) monitoring over one year after GC301 administration. L. Blood GAA activity monitoring over time. In panels A, B, C, and D, the horizontal dashed lines represent the normal physiological thresholds or reference limits for each parameter (e.g., normal LVEF >50%, LVM Z-score between −2 and +2, and VT/kg > 6 mL/kg). Abbreviations: CI, confidence interval; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; GAA, acid alpha-glucosidase. Changes in ventilation efficiency Changes in ventilation efficiency and indices of respiratory muscle weakness were assessed over a 52-week period, and these parameters generally maintained at baseline levels. Tidal volume normalized to body weight (VT/kg), a measure of ventilation efficiency per breath, showed no clear trend of change for any individual patient or for the group median (Fig. 2C). The ratio of time to peak tidal expiratory flow to total expiratory time (tPTEF/tE%), a marker sensitive to expiratory flow limitation and expiratory muscle weakness, also showed no change over one year (p value for the trend in the median tPTEF/tE% across all patients = 0.403) (Fig. 2D). Although patient 2 showed a progressive increase (slope: 0.292; p = 0.047), his values remained the lowest throughout the one-year follow-up. Consistent growth in weight and height following GC301 treatment All tracked patients exhibited a trend of positive weight gain over time. The weight gain slopes for the three boys (patients 1, 2, and 5) were 0.194 kg/month (95% CI: 0.132–0.256), 0.214 kg/month (95% CI: 0.165–0.263), and 0.122 kg/month (95% CI: 0.087–0.157), respectively (Fig. 2E). The slopes for the two girls (patients 3 and 6) were 0.203 kg/month (95% CI: 0.166–0.239) and 0.127 kg/month (95% CI: 0.047–0.206), respectively (Fig. 2F). Patient 4 was not fully documented due to early withdrawal from the study. Nevertheless, the growth trajectories for all patients remained around the 3rd WHO standard percentile line, indicating that their body weight persisted within the range indicative of developmental delay. Similarly, height measurements showed an increase over the observation period. The height gain slopes for patients 1, 2, and 5 (boys) were 0.511 cm/month (95% CI: 0.427–0.595), 0.729 cm/month (95% CI: 0.617–0.841), and 0.558 cm/month (95% CI: 0.259–0.858), respectively (Fig. 2G). The slopes for patients 3 and 6 (girls) were 0.465 cm/month (95% CI: 0.400–0.531) and 0.858 cm/month (95% CI: 0.443–1.274), respectively (Fig. 2H). Despite the overall increase, the height measurements for the boys generally fell around the 3rd WHO standard percentile, and only one girl (patient 6) fell between the 25th and 75th WHO standard percentiles. Motor function outcomes The GMFM-88 scores over 52 weeks revealed divergent patient trajectories (Fig. 2I). Patient 1 (slope: 2.012, p = 0.016) and patient 2 (slope: 2.831, p = 0.007) demonstrated a trend of motor improvement, which is consistent with the observed motor milestones. In contrast, patient 6 showed a steady decline with slope of −0.681. This indicates a heterogeneous individual response in motor function over the 52 weeks post-treatment. All patients exhibited delays or difficulties in achieving developmental milestones to varying degrees. While patients 4, 5, and 6 did not exhibit changes in developmental milestones following GC301 gene therapy, patients 1, 2, and 3 achieved previously unattained milestones such as independent walking during the one-year follow-up period (the independent standing and walking of patient 1 occurred in the 13th month) after treatment (Fig. 2J). Consistent enhancement in quality of life This panel shows the longitudinal trend of the Pediatric Quality of Life Inventory (PedsQL) Neuromuscular Module scores for six patients over 52 weeks. The median score (black line) demonstrates an improving trend with a positive slope of 0.269 (95% CI: 0.097–0.441, p = 0.012) (Fig. 2K). This indicates a potential enhancement in disease-specific health-related quality of life from the patients' or caregivers' perspective during the one-year follow-up period. Blood GAA activity over time Following treatment, all patients experienced an immediate and sharp rise in blood GAA activity, peaking within the first 1–2 weeks. Subsequently, the median of activity level showed a declining trend, with a negative median slope of −0.003 (95% CI: −0.005 to −0.001, p = 0.004), and eventually stabilized at a lower, yet sustained level (Fig. 2L). AAV9 ensures the expression of GAA within skeletal muscle tissue To assess the efficacy of GC301 in delivering the therapeutic gene to skeletal muscle and achieving functional GAA expression, muscle biopsies were obtained from three patients at week 0 and 26 weeks (week 39 for patient 5) post-treatment. Quantitative analysis showed a median vector DNA level of 9030 vg/μg genomic DNA (IQR: 6480–27,100) post-treatment, compared with 5 vg/μg genomic DNA (IQR: 5–5) at baseline (Fig. 3A). The median intramuscular GAA enzyme activity post-treatment was 24.67 nmol/mg/hr (IQR: 8.28–110.09), also higher than the baseline value of 1.24 nmol/mg/hr (IQR: 0.82–2.51) (Fig. 3B). Assessment of AAV9-mediated GAA gene transduction efficiency and muscle histology. This figure evaluates the efficacy of AAV9 in delivering the GAA gene to skeletal muscle cells in patients 3, 5, and 6. Analyses include: (A) direct measurement of vector DNA; (B) intramuscular GAA enzyme activity; (C) intramuscular GAA protein expression; and (D) comprehensive histological evaluation of muscle pathology. The histological assessment comprises: glycogen accumulation visualized by Periodic Acid-Schiff (PAS) staining; muscle fiber integrity and cytoskeletal structure assessed via Desmin immunohistochemistry; mitochondrial function and oxidative capacity indicated by Succinate Dehydrogenase (SDH) activity; and neutral lipid deposition revealed by Oil Red O staining. All microscopic images were taken at 200× magnification. Additionally, panel (E) quantifies glycogen content in muscle tissue before and after GC301 treatment. F, muscle pathology was evaluated semi-quantitatively using Modified Heckmatt ultrasound scores on the quadriceps muscle. Western blot analysis demonstrated a marked increase in GAA protein levels in post-treatment muscle samples from all three patients compared to baseline. Distinct bands corresponding to the precursor (∼110 kDa) and active (∼76-70 kDa) forms of GAA were clearly detected, indicating successful protein production and processing (Fig. 3C). Histopathological assessment of muscle tissue in three participants In the subset of participants who underwent paired biopsies (n = 3), GC301 treatment was associated with apparent improvements in skeletal muscle histopathology. To assess whether restored GAA expression correlated with glycogen clearance, we employed exploratory Periodic Acid-Schiff (PAS) staining, a histological method specific for carbohydrates. Qualitative review revealed prominent glycogen accumulation (abundant pink-stained areas) at baseline. Post-treatment, the intensity of PAS-positive staining appeared reduced, suggesting decreased glycogen deposition. Quantitative analysis in these three patients (Patients 3, 5, and 6) supported this observation, with mean PAS intensity values decreasing from 296.1 ± 203.8 to 50.8 ± 74.9 for patient 3, 269.5 ± 223.6 to 85.4 ± 101.8 for patient 5, and 169.5 ± 111.0 to 46.2 ± 73.1 for patient 6 (Fig. 3E). Desmin staining, evaluated qualitatively, revealed extensive disorganization of muscle architecture at baseline. Following treatment, Desmin expression exhibited a more uniform pattern, suggesting a potential restoration of muscle fiber integrity. Similarly, Succinate Dehydrogenase (SDH) staining suggested enhanced mitochondrial activity, with increased density and uniformity of blue granules in post-treatment samples. Oil Red O staining showed a trend toward reduced lipid droplet accumulation post-treatment (Fig. 3D). Finally, the modified Heckmatt ultrasound score showed a minimal downward trend over 52 weeks (median slope −0.015; p = 0.512; Fig. 3F). Given the known limitations of the Heckmatt scale, these findings should be interpreted with caution.30 Changes in enzymological indicators related to muscle pathology Although the serum levels of CK, LDH, AST, and ALT were originally intended to monitor potential tissue damage from gene therapy, their dynamic profiles have nonetheless enriched our understanding of IOPD pathology from an enzymatic perspective. As shown in Fig. 4, all four enzyme levels remained above the normal range throughout the entire observation period in all six patients. This indicates the presence of chronic and sustained systemic muscle and liver injury, even after GC301 treatment (pre-enrollment data are not shown). Of the six patients monitored, three (patients 3, 5, and 6) exhibited increasing trends in CK levels over the one-year observation period, with individual slopes of 1.919, 1.495, and 2.841, respectively. Furthermore, the median CK level across all patients also showed an increasing trend, with a slope of 1.737 (95% CI: 0.481–2.994) (Fig. 4A). The median LDH level across all patients showed an increasing trend, with a slope of 0.602 (95% CI: 0.304–0.900) (Fig. 4B). A one-year follow-up of patient 2 revealed a decreasing trend in AST levels, with a slope of −0.192 (95% CI: −0.383 to −0.001). In contrast, the median trend across all patients was not statistically significant (Fig. 4C). Although all patients had ALT levels exceeding 40 U/L, no discernible trend of change was observed over the one-year observation period (Fig. 4D). Dynamic changes in clinical enzymatic indicators related to muscle pathology. The dynamic changes in the levels of creatine kinase (A), lactate dehydrogenase (B), aspartate aminotransferase (C), alanine aminotransferase (D), triglyceride (E), and total cholesterol (F) were monitored. The horizontal dashed lines in each panel indicate the upper limit of the normal reference range for the respective serum enzymes: 40 U/L for both alanine aminotransferase (D) and aspartate aminotransferase (C); 200 U/L for creatine kinase (A); 250 U/L for lactate dehydrogenase (B); 1.7 mmol/L for triglyceride (E); and 5.7 mmol/L for total cholesterol (F). In addition to revealing these centralized trends, clinical enzymology also indicates that unexpected events during treatment can cause fluctuations in these markers. For instance, conditions such as pneumonia and upper respiratory tract infections may lead to transient elevations in these enzyme levels (Fig. 4B–D). Host immune response to AAV9 and GAA Consistent with our previous findings,2 all six patients developed sustained AAV-binding IgG after GC301 administration (Supplementary Table S2). T-cell immune responses were transient and asymptomatic. Mild T-cell responses to AAV9 capsid were detected in two patients (P2 and P6), while responses to GAA-derived peptide pools were observed in three patients (P1, P2, and P5); none of these cellular responses were associated with elevated liver enzyme levels, or clinical deterioration (Supplementary Table S3). Discussion Since ERT became the standard treatment for IOPD, most patients in real-world settings have required prolonged, bi-weekly infusions.31 Even with sustained, high-dose ERT, however, limited tissue distribution of the drug remains a recognized challenge.32,33 Our previous study showed that a single intravenous dose of GC301 enabled infants under 1 year of age with minimal prior ERT to remain ventilator- and ERT-free for at least one year.2 Whether this therapeutic strategy can also sustain survival, growth, and development in older children with IOPD after extended, high-dose ERT remains unknown. This trial enrolled six patients with IOPD with a median age of 25.5 months, a median of 30 prior ERT sessions, and a median ERT duration of 19.5 months. Patient selection criteria were meticulously balanced: individuals with end-stage cardiorespiratory failure were excluded to avoid ethical risks associated with rapid disease progression upon ERT withdrawal, as were those with normal motor development, to ensure post-treatment clinical improvements could be effectively assessed. Furthermore, although the pre-treatment ERT washout period varied (4–20 days), this variance in this phase is unlikely to meaningfully interfere with the 52-week long-term therapeutic outcomes, given the extremely short half-life of exogenous rhGAA. A one-year observation period following a single infusion of GC301 at 1.2 × 1014 vg/kg revealed that: (1) High safety profile, with the incidence of grade ≥3 adverse events being 2.31 person-time, possibly drug-related adverse events were 5.58 person-time; (2) 100% ventilator-free survival rate, except for patient 4 who required ventilator support once on day 60 post-treatment due to pneumonia, while pulmonary function parameters (VT/kg and tPTEF/tE) generally showed no overall improvement in the six patients (except tPTEF/tE of patient 2); (3) Patients' cardiac function and structure remained at baseline levels. The overall trend of LVEF showed a slight declining trend and the overall trend of LVM Z-score showed a slight upward trend, however, the median trends for both indicators are very weak, with slopes of −0.013 and 0.004, respectively; (4) All documented children exhibited growth during the observation period, although the gene therapy did not improve their percentile distribution in growth parameters; (5) Patients 1, 2, and 3 achieved key milestones such as unassisted crawling, standing, and walking during the observation period (the independent standing and walking of patient 1 occurred in the 13th month); (6) Muscle biopsy findings from the subset of three treated participants suggested reduced skeletal muscle glycogen accumulation, alongside apparent improvements in muscle structure; (7) PedsQL 3.0 NMM scores indicated overall improvement in quality of life; and (8) Improvement in GMFM-88 scores was observed in patients 1 and 2. The longitudinal clinical trajectories of the six patients were reviewed from the initial diagnosis through the ERT period to the 52-week observation period following gene therapy (Supplementary Figure S1). At initial diagnosis, these patients presented with myocardial hypertrophy. Following ERT, the hypertrophy was ameliorated to a certain extent in all patients. Upon enrollment, the LVM Z-score showed a slight increase in 4 of the 5 pediatric patients, while their LVEF was maintained within the normal range. Five out of six patients displayed no improvement in pulmonary function; however, all patients remained alive and ventilator-free over the 52 weeks (with the exception of patient 4, who developed severe pneumonia). Considering that the natural history of IOPD indicates a median age of death of only 8.7 months in untreated patients, predominantly due to cardiopulmonary failure,34 the outcomes following GC301 treatment provide preliminary evidence of clinical activity in this study population. The gross motor function is influenced by multiple factors, such as the patient's baseline status at the time of treatment and cardiopulmonary function. For instance, the decline in motor function relative to baseline observed in patient 6 at 52 weeks post-treatment must be interpreted in light of the fact that this patient presented with severe motor impairment upon enrollment (a GMFM-88 score of only 15.8%). Such an advanced underlying disease state suggests that chronic glycogen accumulation had likely already caused irreversible muscle fiber damage, atrophy, and fibrofatty replacement. Despite histological evidence of reduced glycogen accumulation at the cellular level, residual muscle was insufficient to reverse the already established clinical motor deficits. Thus, improvement in gross motor function may require a longer observation period. Similarly, given that alteration of LVM is a slow pathophysiological process, a 52-week observation window may be insufficient to fully capture long-term structural changes. An ongoing long-term follow-up study is concurrently being conducted to evaluate these patients for up to five years post-treatment, and subsequent results will be reported in due course. Although all three patients had received 19 to 85 rounds of ERT prior to GC301 gene therapy, baseline muscle histopathological analysis revealed substantial glycogen accumulation, numerous lipid vacuoles, compromised structural integrity, uneven mitochondrial distribution, and fatty infiltration. After GC301 gene therapy, histological data from muscle biopsies demonstrated marked changes: glycogen accumulation was reduced, muscle integrity was enhanced, mitochondrial density increased and became more uniformly distributed, and fatty infiltration was reduced. These findings suggest that GAA gene therapy using a recombinant AAV9 vector is associated with increased GAA expression in muscle tissue and subsequent alterations in muscle structure and function. In contrast, changes in respiratory muscle-related indicators were less apparent, which may be attributed to the distribution characteristics of the intravenously administered AAV9.35,36 GAA enzyme activity was monitored in peripheral blood for one year and in muscle tissue at 26/39 weeks post-gene therapy. Patient 1's baseline GAA activity was 82.8 μmol/L/hr, likely reflecting residual effects from prior ERT. The median blood GAA activity in the six patients gradually declined from 1.9 μmol/L/hr at week 1 to 0.9 μmol/L/hr by week 52. The peak in blood GAA activity observed during the initial phase (week 4–13) of gene therapy likely originates in part from AAV9-mediated GAA gene expression in circulating peripheral blood cells and in part from minor secretion by the transduced tissues. As peripheral blood cells have a relatively short lifespan and undergo rapid natural turnover and apoptosis, the subsequent decline in blood GAA levels is biologically expected. In contrast, the median GAA activity in muscle tissue at week 26/39 was 24.67 nmol/mg/hr. If the enzymatic activity of muscle tissue is simply converted to the same unit as that of serum enzyme activity, assuming 1 mg of muscle is equivalent to 1 mL in volume, the enzymatic activity in muscle tissue at 26/39 weeks after gene therapy was 24.67 μmol/L/hr, which is substantially higher than the corresponding serum enzyme activity at similar time points (1.0 μmol/L/hr at week 26 and 0.7 μmol/L/hr at week 39). While the standardization of GAA enzyme activity measurement and its sensitivity and specificity in diagnosing Pompe disease are still in the process of being improved,37,38 based on the patients' clinical manifestations over the one-year observation period, our gene therapy dosage appears to maintain effective GAA levels for at least one year. Throughout the 52-week observation period, CK, LDH, AST, and ALT remained generally elevated, with CK showing an overall upward trend. These biomarkers were persistently elevated from diagnosis through the ERT period and generally remained near baseline levels after gene therapy (Supplementary Figure S1). The fluctuation and upward trend in CK levels may be driven by a transient decrease during initial prophylactic corticosteroid treatment, followed by a gradual rebound as steroids were tapered.39 In IOPD, massive glycogen accumulation causes lysosomal swelling and rupture, releasing hydrolytic enzymes into the cytoplasm and triggering cell necrosis, pyroptosis, and apoptosis, ultimately driving chronic sterile inflammation—a possible key mechanism underlying elevated CK, LDH, AST, and ALT.40 However, CK is also a key enzyme in creatine metabolism. It is possible that GAA deficiency disrupts cellular energy homeostasis, which prompts skeletal muscle to upregulate creatine metabolism for energy supply. Thus, metabolic stress and pathway switching, rather than active muscle destruction, are the likely core drivers of persistently elevated CK in Pompe disease.41, 42, 43, 44 Currently, no evidence links changes in CK levels to therapeutic efficacy. Although next-generation ERT has been optimized, it still relies on lifelong, bi-weekly exogenous enzyme infusions.45,46 In contrast, AAV9-mediated gene therapy achieves sustained, multi-organ endogenous GAA expression with a single dose. Recent preclinical studies47, 48, 49 have shown that muscle-targeted promoters can lower vector doses while maintaining high expression in skeletal and cardiac muscles. However, IOPD is a severe multisystemic disorder requiring widespread and sustained functional improvement across multiple organs, including the myocardium, skeletal muscle, respiratory muscles, and the CNS.46 To meet this therapeutic goal, the GC301 expression cassette uses a ubiquitous CMV enhancer/chicken β-actin promoter to drive broad gene expression, achieving high GAA activity not only in muscles but also in the liver. It is theoretically possible that the liver could act as a systemic reservoir, secreting functional GAA into the circulation. However, the relatively low and declining blood GAA levels observed in our study suggest that the contribution of liver-derived GAA to muscle glycogen clearance is likely limited. Therefore, the observed changes in skeletal muscle are likely associated with direct transduction of muscle fibers by GC301, as supported by the higher GAA enzyme activity and protein expression detected in muscle biopsies. Regarding safety, GC301 treatment in patients with IOPD compares favorably to high-dose AAV trials for other neuromuscular diseases, which have shown severe toxicity such as hepatotoxicity or thrombotic microangiopathy.50, 51, 52 In our cohort, post-dosing ALT and AST elevations were transient and self-limiting. This favorable safety profile is largely due to our prophylactic corticosteroid regimen and the baculovirus-insect cell system.24 This system utilizes biologically contained, species-specific expression cassettes and entirely lacks the AAV p5 promoter.53 These design features are intended to minimize the risk of generating replication-competent AAV (rcAAV) by limiting the homology sequences required for recombination, thereby providing a built-in safety mechanism at the point of delivery.53,54 In contrast, the HEK293 triple-transfection system has an inherent biological potential to generate rcAAV through recombination.55,56 In a reported case, metagenomic analysis of a liver biopsy from a patient with severe hepatitis identified residual sequences from manufacturing plasmids and complex vector genome rearrangements. The authors postulated that these contaminants, potentially in conjunction with helper viruses like HHV-6B, might contribute to hepatotoxicity, although a formal causal link remains to be established.57 The hypothesis is that in a patient with a high vector load and latent human herpesvirus 6B (which may provide helper functions), these rcAAV particles find conditions conducive to amplification. Serious adverse events were mainly upper respiratory tract infections and pneumonia, which reflect the underlying disease characteristics due to inherent respiratory muscle weakness and impaired airway clearance in IOPD. Apart from expected steroid-induced dyslipidemia, drug-related adverse events were confined to mild post-dosing fever and transient transaminase elevations. This study has several limitations. First, the small sample size was necessitated by the rare and specific clinical progression of Pompe disease, which also precluded a randomized controlled trial design; therefore, all statistical results should be interpreted with caution. Second, as a single-center investigation, the study's representativeness and external validity are limited.58 Third, this study employed only a single dose and was unable to determine the optimal therapeutic dose. Furthermore, this study is subject to two key limitations regarding muscle biopsy analyses. First, the limited tissue yield was entirely allocated to pre-specified core assays, precluding quantification of glycogen content, autophagic flux, and the prevalence of central nuclei. Second, although three participants (Patients 1–3) achieved major motor milestones—including independent walking—paired pre- and post-treatment biopsies were available only for patient 3. Consequently, evidence correlating gene therapy with molecular histopathological improvement remains limited. Finally, interpretation of neurological outcomes is constrained by two factors: first, the age-dependent permeability of the blood–brain barrier to AAV9,59 which may have limited CNS transduction in our cohort (median age 25.5 months), although preclinical data in adult and aged mice suggest that partial CNS activity persists; and second, the 52-week observation period, which may be insufficient to detect slowly progressive neurological effects. Recombinant AAV9-mediated GAA gene therapy shows promise in decreasing the substantial reliance of IOPD patients on ERT, while also inducing changes in both development and skeletal muscle pathology. Contributors Guang Yang and Xiaobing Wu, were responsible for study design, clinical trial organization and implementation, and overall supervision. Xiaobing Wu, Wenhao Ma, and Xiaoyan Dong were responsible for the development, manufacturing, and quality control of the gene therapy product(s). Shuangqing Yu and Xiaodong Wang were responsible for the study protocol and project administration. Xinting Liu and Linyan Hu were responsible for the day-to-day conduct of the clinical trial, data collection, and patient follow-up. Yingying Mao and Wen He were responsible for clinical research logistical coordination. Xueyuan Guo, Min Li, Yiru Wang, Yun Yuan, Zhaoxia Wang, Shu Zhang, and Shiwen Wu were responsible for multi-department clinical consultation and associated medical examinations. Yue Wang, Xinting Liu, and Guang Yang were responsible for formal analysis and validation. Xinting Liu, Linyan Hu, and Guang Yang — all MDs affiliated with the Senior Department of Pediatrics, Chinese PLA General Hospital, and Department of Pediatrics, First Medical Center, Chinese PLA General Hospital — had full and direct access to the raw, participant-level data, including all clinical measurements and laboratory results. Xinting Liu, Yue Wang, and Guang Yang were responsible for visualization, and for writing the original draft and reviewing and editing the manuscript. All authors have read and approved the final paper. Data sharing statement De-identified individual participant data underpinning the results reported in this article, along with the study protocol, statistical analysis plan, and informed consent form, will be made available indefinitely upon publication. Researchers may request access by contacting the corresponding author. Access will be granted to researchers with a methodologically sound proposal, following review and approval by a designated review committee and execution of a data access agreement. Data are provided for approved analyses, such as individual participant data meta-analysis, without investigator support. Declaration of interests None exist. Acknowledgements This Project is supported by the National Key R&D Program of China (Grant No. 2023YFC3403300) and the Beijing Science and Technology Plan Project (Grant Z231100004823022). We extend our sincere gratitude to all participants in this clinical trial, including the children and their parents. We also thank the entire medical and nursing staff of the Department of Pediatrics at the First Medical Center of the Chinese PLA General Hospital for their professional clinical care. Special thanks are due to Professor Yan Meng (Guangzhou Women and Children's Medical Center) and Professor Qiang Shi (Department of Neurology, First Medical Center of the Chinese PLA General Hospital) for their contributions during the initial phase of the study. Footnotes Supplementary data related to this article can be found at https://doi.org/10.1016/j.eclinm.2026.104081. Contributor Information Xiaobing Wu, Email: wuxiaobing@bj-genecradle.com. Guang Yang, Email: yangg301@126.com. Supplementary data Supplementary Figure and Tables References Associated Data Supplementary Materials Supplementary Figure and Tables