Clinical manifestations

Presentation and course

The advent of enzyme replacement therapy marks the development of a shifting natural history, including new phenotypic manifestations, disease complications, and understanding of the spectrum of Pompe disease (122). Presentation continues to diversify by age at onset, extent of organ involvement, and degree of myopathy. The severity of disease manifestations tends to reflect the level of residual enzyme activity, with lower enzymatic activity typically correlating to greater disease burden (121; 131). In addition to the underlying genetic variants, a number of additional factors affecting clinical presentation are being identified, including the environment, diet, exercise, and other modifying genetic elements. Genetic modifier such as the angiotensin-converting enzyme (ACE) variants with genotype DD (deletion/deletion), for example, has been linked to an earlier age of onset, faster rate of progression, and more severe muscle pain (49; 50). Patients with the DD genotype appeared to have a reduced response to enzyme replacement therapy, when compared to patients with II or ID (10). However, recently, irrespective of the angiotensin-converting enzyme variants, a large variation was observed in the disease severity and enzyme replacement therapy response in patients with the most common genetic variant c.-32-13T>G in late-onset Pompe disease (LOPD) in a European cohort (136). Another genetic modifier is the ACTN3 gene; a null polymorphism in the ACTN3 gene was shown to have a positive effect on treatment efficacy with ERT (198). Female sex has been identified as a good prognostic factor for the effect of ERT on muscle strength. Another important factor that can be used to predict clinical outcomes in patients with Pompe disease is the cross-reactive immune material (CRIM) status. CRIM-positive patients have detectable GAA protein, whereas CRIM-negative status denotes absence of GAA protein, and these patients tend to have poorer outcomes compared to CRIM-positive patients (120) (See Section 4).

Infantile-onset Pompe disease (IOPD). Classic IOPD is the most severe form of GAA deficiency, with enzyme activity levels less than 1% of normal in skin fibroblasts. These infants manifest in the first days to weeks of life, with cardiomyopathy and features that can be subtle such as feeding difficulties, cardiac arrhythmias (ie, supraventricular tachycardia), increase in creatine kinase (30), aspartate aminotransferase (113), alanine aminotransferase (197), and urinary glucose tetrasaccharide (Glc4, also referred to as hexose tetrasaccharide, Hex4), to the more florid findings of diffuse hypotonia and weakness, giving them a "rag doll" appearance (floppy baby syndrome). Hypertrophic cardiomyopathy, massive cardiomegaly, and diffuse skeletal myopathy are hallmarks of classic infantile Pompe disease. Muscle bulk is often increased secondary to glycogen deposition, especially in the gastrocnemius muscles. Macroglossia is common. Despite their extreme weakness, these infants are usually alert and interested in their environment. Obstructive sleep apnea and hypoventilation are commonly noted in patients with IOPD, even in those that do not have symptoms of sleep-disordered breathing. Without treatment, infants with classic infantile Pompe disease have feeding and swallowing difficulties and failure to thrive, and patients usually succumb to the complications of progressive cardiac/muscle weakness – increased rate of pulmonary infections, cardiac failure, or respiratory failure – before 2 years of age.

Enzyme replacement therapy has increased life expectancy and subsequently altered the natural history of the disease, including new phenotypic manifestations and disease complications. In the United States, Prater and colleagues described the emerging phenotype of patients with IOPD on enzyme replacement therapy by taking a retrospective look at 11 patients who were ventilator free, started treatment prior to 6 months of age, and survived past their fifth birthdays (188). All patients were an average of 8 years old and free of invasive ventilator support at the time of the latest assessment. In general, the cohort showed improvement in cardiac and gross motor function. However, they continued to present with some residual muscle weakness, hypernasal speech, hearing loss, risk for arrhythmias, dysphagia with aspiration risk, and osteopenia. Other emerging issues of patients with IOPD on long-term enzyme replacement therapy include poor anal-sphincter tone and the potential for basilar artery aneurysms (180; 226). Clinical manifestations of sensorineural hearing loss and swallowing dysfunction in patients with IOPD have been attributed in part to bulbar weakness, despite being treated with enzyme replacement therapy (231; 196).

Autopsy findings in patients with IOPD have revealed glycogen accumulation in anterior horn cells of the spinal cord and neurons of brainstem nuclei, as well as neurons of the cerebral cortex to a lesser extent (77; 17). To better characterize these abnormalities in patients with IOPD on enzyme replacement therapy, McIntosh and colleagues studied brain imaging (including CT and/or MRI) from 23 patients with IOPD (17 CRIM-positive, 6 CRIM-negative) (159). In this study, 16 patients (70%) had neuroimaging abnormalities consisting of ventricular enlargement (VE) and/or extra-axial cerebrospinal fluid accumulation (EACSF) at baseline, with delayed myelination in 2 patients. Among the 8 patients who received follow-up assessments, 7 demonstrated normalization of ventricular enlargement and extra-axial cerebrospinal fluid accumulation. White matter changes were noted in 2 of 3 patients imaged after 10 years (159). A study systematically quantified the white matter abnormalities seen in 10 out of 12 children with IOPD who were on enzyme replacement therapy, using a grading system on MRI brain (128). Variations were seen in the severity of white matter involvement; however, there were certain areas that had an early involvement among all the children. In the brain MRIs from survivors with IOPD, periventricular and subcortical areas were involved earlier and more commonly than the deeper white matter areas (62; 127; 128). At this time, further investigation is warranted to determine the cause of these white matter abnormalities in the MRI and to understand their impact on cognition and development.

The short life expectancy of patients with IOPD prior to the advent of enzyme replacement therapy once precluded studies on cognitive functioning in these patients but no longer poses an obstacle. With the evidence of the central nervous system involvement, its impact on developmental outcomes becomes another growing concern. Spiridigliozzi and colleagues found that patients with classic IOPD 18 months old and under did not exhibit any cognitive impairment (219). A study on the long-term progression of cognitive ability in infantile Pompe patients on ERT showed that those with classic infantile Pompe disease maintained a lower than average IQ, whereas those with nonclassic Pompe disease had a higher IQ than their average peers that improved over time (220). This finding of cognitive impairment in classic infantile Pompe disease was found in another cohort, and impaired motor function was cited as a complication in assessing cognitive ability in children less than 5 years of age (61). Cognitive decline with moderate intellectual disability was reported in a patient with classic infantile Pompe disease (63). Using an array of verbal and nonverbal tests, a study found that long-term survivors of IOPD patients (range = 5 years, 6 months through 17 years of age) had a median IQ of 84 on Wechsler scales and of 92 on Leiter scales. The study highlights the importance of using appropriate tests to measure verbal and nonverbal abilities (based on each individual’s motor skills, speech and language abilities, auditory function, and native language), so as to enable an accurate assessment of whether there is a learning disability or an intellectual disability (221). Developmental outcomes from 12 children with IOPD were assessed simultaneously during their MRI brain assessments (128). The study showed that children with IOPD had relative weaknesses in processing speed, fluid reasoning, visual perception, and receptive vocabulary. In addition, however, due to the small sample size, relationships between brain MRI and cognitive IQ tests did not yield any significant findings. To better understand the impact on behavior, emotions, and social aspects of children with Pompe disease, Korlimarla and colleagues further showed that most of the children with Pompe disease exhibited age-appropriate behavior and emotional functioning, compared to age-matched and gender-matched peers (129). The findings were based on parent reports of 21 children with Pompe disease (n=17 IOPD; 5- to 18-year-old children). However, there were certain domains such as negative mood symptoms, learning problems, decreased participation in social activities, and attentional difficulties, which were more frequently reported in children with IOPD in comparison to same-aged peers. In addition, the study showed that despite the physical challenges of a chronic illness, and increased absenteeism due to weekly/biweekly infusions, most of the children also attended regular classrooms, along with special accommodations to help them.

Late-onset Pompe disease (LOPD). In contrast to infantile-onset, which has less than 1% GAA enzyme activity, most patients with LOPD have about 1% to 40% of residual GAA enzyme activity in skin fibroblasts (Section 4). Disease onset ranges from infancy to as late as the sixth decade of life (132; 196; 91). Given the typically higher percentage of enzyme activity in LOPD, muscular weakness tends to be less severe, and therefore, the disease onset is slower, and disease progression is typically less aggressive than in IOPD.

Common presentations in the first year of life include swallowing difficulty, feeding difficulty, and sleep apnea. In LOPD, sleep-disordered breathing is common and comprises both hypoventilation and obstructive sleep apnea. Noninvasive ventilation significantly improves respiration and sleep quality (25). However, as the disease progresses, in some instances noninvasive ventilation becomes insufficient to maintain adequate pulmonary ventilation, and patients require continuous ventilator support such as a tracheostomy or a tube inserted into the upper airway (156). In addition to presenting with limb girdle and respiratory involvement, patients can present with ptosis, rigid spine syndrome, lingual weakness, and vascular complications such as basilar artery aneurysms. Children with LOPD may exhibit delayed motor milestones.

A study that utilized data from the Pompe Registry, the largest Pompe database in the world with information from 742 patients in 28 countries, found that out of 517 patients with LOPD, 89.2% experienced proximal muscle weakness in the lower extremities, 72.9% experienced proximal muscle weakness in the upper extremities, and 65.2% experienced muscle weakness in the trunk (32). In terms of respiratory manifestations, 65.6% experienced shortness of breath after exercise, and about 8.7% had used respiratory support at some point. A 2-year study in 52 Dutch patients with LOPD, ranging in age from 4 to 81 years, showed a broad spectrum of disease severity (81). Over 2 years, there was a significant increase in the number of daily hours of respiratory support and a significant decrease in functional activities, such as riding a bicycle. Other motor activities declined less markedly and some did not change, possibly because they were severely compromised at baseline. In another study, a group of 58 late-onset patients showed increased weakness in both the lower and upper extremities as well as the respiratory muscles over a 12-month period (249). In a study of 225 patients, respiratory failure was the most common cause of death among nonclassic patients with IOPD and patients with LOPD, and it was present in about 72% of patients in this study (248).

Weakness predominates in truncal and proximal muscles (with lower extremities more involved than upper extremities). Clinically significant differences in gait, posture, selective muscle weakness, and contractures in muscle groups in patients with LOPD, when compared to their IOPD peers, have been extensively reviewed (34). Quantitative MRI data from patients with LOPD have also revealed significant fatty infiltration of musculature, largely correlating with muscle weakness identified via clinical muscle function assessments; additionally, evidence suggests that MRI data are able to monitor progression of muscle fatty infiltration over time, and in some cases, they have been shown to be more sensitive than traditional disease outcome measures such as physical examination, muscle function tests, or patient-reported outcomes (199; 33; 103; 71; 238; 149; 117). Nerve conduction abnormalities are also noted, and myotonic discharges may be prevalent and abnormal (American Association of Neuromuscular & Electrodiagnostic Medicine 2009). However, it is unclear whether these abnormalities occur due to damage in a muscle, a nerve, or both.

Respiratory muscles are also affected in LOPD. Respiratory insufficiency may be the presenting complaint usually associated with daytime sleepiness, fatigue, morning headache, exertional dyspnea, or sleep-disordered breathing (26; 156). Diaphragmatic weakness is a significant feature in Pompe disease and is the major cause of sleep-disordered breathing and early respiratory failure in patients with Pompe disease (Wens et al 2015; 215). Muscle MRI data have shown that respiratory insufficiency in LOPD is mainly due to diaphragmatic weakness with relative sparing of intercostal muscles (76). A decrease in forced vital capacity (FVC) from the upright to supine position is characteristic in LOPD and highlights the involvement of the diaphragm (32). It is important to note that respiratory insufficiency is not always commensurate with limb weakness, so respiratory function should be monitored independently to allow timely implementation of mechanical ventilation or tracheostomy (181). In LOPD, respiratory failure is usually the cause of death.

Although the heart is not extensively involved in LOPD like in IOPD, Wolff-Parkinson-White (WPW) syndrome, sinus arrhythmias, atrio-ventricular blocks, and other electrocardiographic abnormalities have been seen in late-onset Pompe patients (Sancconi et al 2014; 37). Left ventricular hypertrophy (LVH) has been reported now that the phenotype is being better documented and further understood (168; 72; 142; 08). Thirteen percent (9 out of 64) of patients with LOPD showed variable degree of hypertrophy (LVH and septal), and there was reversal of the cardiac hypertrophy with enzyme replacement therapy (for 3 years) in 1 of these 9 patients (05). It was shown that patients with LOPD with the c.-32-13T>G variant (IVS1 splice site variant) do not have severe cardiomyopathy; however, they can present with arrhythmias and aortic root dilatation in adulthood (92). In contrast, patients without the c.-32-13T>G variant can present with cardiomyopathy (92). Increased glycogen accumulation in arterial walls has been considered responsible for the observed increased aortic stiffness and blood pressure, increasing the potential risk of cardiovascular diseases in adult patients (173; 245). Dilated arteriopathy involving the ascending aorta has been confirmed as a complication (65). These cardiac findings are blurring the lines of what was once the definitive feature of classic IOPD and reinforcing the idea that Pompe disease is, in fact, a continuous spectrum. Other than the blood vessels of the heart, there is strong evidence of glycogen accumulation in the cerebral arteries. Glycogen accumulation in the smooth muscles of the cerebral arteries causes intracranial aneurysms that can lead to early death in LOPD (151; 155; 130; 139; 204).

Involvement of central/peripheral nervous system has also been reported in late-onset children and adults with LOPD. A study revealed the prevalence of hearing loss in patients with LOPD slightly exceeded the normative data of the general population, with both sensorineural and conductive hearing loss observed in this cohort of 11 patients with LOPD (90). In another retrospective study, 3 of 12 patients with LOPD presented with oropharyngeal dysphagia as a manifestation of bulbar weakness. Small fiber neuropathy, which often presents with painful paresthesias in the extremities or autonomic dysfunction, was reported in a study by Hobson-Webb and colleagues (98). The study used the Small Fiber Neuropathy Screening List, a 21 question screening tool with a high sensitivity for detecting small fiber neuropathy. Fifty percent of the 44 patients in the study had a mean score of more than 11, indicating that they are at risk for small fiber neuropathy. Small fiber involvement could be due to the deposition of glycogen in Schwann cells of peripheral nerves (99). Musumeci and colleagues reported neuroimaging findings from 21 patients with LOPD, including presence of white matter lesions (Fazekas total score > 2) in 67% of patients, signs of cerebral vasculopathy in 12 of 21 patients (57%), dolichoectasia of the vertebrobasilar system in 11 of 21 patients (52%), and intracranial aneurysm in 3 of 21 patients (14%) (170). Functional MRI revealed decreased brain connectivity in the salience network, mild impairment in executive functions, and gray matter atrophy correlating with age and disease duration. On the other hand, upon comparison between white matter lesions identified via MRI in 19 treated patients with LOPD and those seen in 38 matched controls, Schneider and colleagues concluded that there was no significant difference between the 2 groups based on frequency of white matter lesions, Fazekas grade, and white matter volume (206). Although cognitive function is largely thought to be unaffected by LOPD, impaired intelligence was noted in one adult patient based on the Wechsler Adult Intelligence Scale (WAIS-III) (169). Korlimarla and colleagues reviewed brain MRI and neurologic outcomes in patients with LOPD available in the literature, and they found that 31 adult patients with LOPD showed mild impairment in executive functions and visual-constructive abilities (28; 127; 170). In another study, Korlimarla and colleagues showed that 2 children with LOPD did not have any abnormalities in the brain MRI and were typically developing when compared to their peers without Pompe disease (128). However, while focusing on behavioral aspects, there were some relative weak domains such as negative mood symptoms and peer relations (behavior as endorsed by their parents) for three fourths of children with LOPD (129). Taken together, these findings warrant further investigation into the impact of LOPD on the central nervous system and the potential effects on cognitive function and behavior. It must be remembered that the data at this time are sparse, and other causes of cognitive decline have not been excluded or explored.

In addition to involvement of muscles, heart, and nervous system, many other body organs have been found to be affected in LOPD (37). A study involving 35 patients with LOPD looked at the prevalence of lower urinary tract symptoms and bowel involvement. The individuals in the study, which included both men and women, reported urinary incontinence, dribbling, weak urine stream, post-void dribbling, and inability to stop urination midstream. Bowel incontinence was reported in 45% of patients. The study found a significant association between urinary symptoms and lower extremity function scores and duration of enzyme replacement therapy (160). Involvement of the gastrointestinal system is common in adult patients with LOPD; however, the symptoms are often misdiagnosed or underreported. Symptoms such as swallowing difficulty, chronic diarrhea, postprandial bloating, abdominal pain, and irritable bowel syndrome were reported in patients with LOPD in one study (20). In addition, problems in the skeletal system are also common (37). These include osteopenia/osteoporosis, kyphosis, scoliosis, chest wall abnormalities, and winging scapulae. Gait and posture are commonly affected (34). A study of 22 patients found a high prevalence of asymptomatic vertebral fractures. The fracture prevalence was not found to be associated with muscular and respiratory function and genotype (22).

Prognosis and complications

Until 2006, prognosis was dismal for infantile Pompe disease, which was invariably fatal before 2 years of age. A retrospective, multinational, and multicenter study of 168 patients defined the natural history of untreated infantile-onset Pompe disease (IOPD) (119). Median ages at onset, diagnosis, ventilator dependency, and death were 2 months, 4.7 months, 5.9 months, and 8.7 months, respectively. In another study, 20 untreated patients with infantile-onset from various Dutch care centers and 133 patients from the literature revealed a median age of death of 7.7 and 6 months, respectively (234). Although enzyme replacement therapy has changed the prognosis in some infants and increased the life expectancy of most of them, long-term issues in the surviving infants are emerging and include residual muscle weakness, relative weaknesses of certain domains in cognition and behavior, white matter hyperintense foci (as seen in MRIs), hearing loss, osteopenia, dysphagia with risk for aspiration, risk of arrhythmias, bladder and bowel sphincter incontinence, premature pubarche, ophthalmological issues such as myopia, astigmatism and ptosis, and hypernasal speech (188; 160; 227; 221; 62; 128; 129).

Although it is largely unknown whether late-onset Pompe disease (LOPD) shortens the life span, a study of 268 adult patients with LOPD found that 34 of 268 patients were deceased over a follow-up period of 7 years, of which 23 patients died prior to enzyme replacement therapy (79). The estimated median survival was 27 years after diagnosis. Without treatment, patients with LOPD tend to have a higher mortality rate than the general population, with a median age of death at 55 years (range 23 to 77 years) (79). In a subgroup of 99 Dutch patients, the number of deaths was higher than the expected number of deaths in the general population. Additionally, patients with wheelchair and/or respiratory support had a shorter life expectancy compared to those without wheelchair and/or respiratory support. The cause of death in patients with LOPD is typically attributed to respiratory complications; however, studies have shown that basilar artery aneurysms are also responsible (82; 162; 139). A systematic review and metaanalysis conducted by Schoser and colleagues showed that with enzyme replacement therapy, patients with LOPD had 5-fold lower mortality rate than untreated patients (207). In addition, patients with LOPD on enzyme replacement therapy had remarkable improvement in ambulatory and ventilation status (207). The natural history of Pompe disease continues to change and develop since the advent of enzyme replacement therapy and supportive therapies.

Clinical vignette

A 31-year-old man presented with pneumonia, at which time considerable wasting of proximal limb and trunk muscles was noted. He also reported difficulty with swallowing food. The patient had noted steadily progressive muscle weakness since the age of 24, but he had been able to work and had sought no medical advice except for low back pain (a myelogram was normal). In childhood, he reported no issues and was very active. His 2 brothers and both parents were normal.

Neurologic examination showed marked atrophy and commensurate weakness of the pelvic girdle and shoulder muscles. Both primary and secondary respiratory muscles were weak. Weakness of trunk muscles was associated with thoracic scoliosis and marked lumbar lordosis. Distal limb muscles and craniobulbar musculature were spared. Tendon reflexes were normal. There were no sensory deficits. Mentation was normal.

Six months later, the patient developed respiratory insufficiency with arterial hypoxemia and respiratory acidosis, necessitating permanent tracheostomy. Clinical studies were performed, including lung function tests, an electromyography (EMG), and a punch muscle biopsy. A pulmonary function test showed a decrease in forced vital capacity (FVC) both in the upright and supine position, with greater postural drop of FVC in the supine position. Needle EMG showed abundant spontaneous fibrillations as well as synchronized fibrillations from many proximal and distal muscles. On maximal effort, there was reduced interference. Muscle biopsy, however, was normal.

Laboratory tests showed increased serum creatine kinase (159 IU; normal, less than 45) (30). Urinary glucose tetrasaccharide (Glc₄) levels and serum aspartate aminotransferase (113), alanine aminotransferase (197), and lactate dehydrogenase levels were elevated. GAA activity in dried blood spots showed deficiency. GAA sequencing revealed compound heterozygosity for 2 pathogenic variants, c.-32-13T>G (IVS1-13T> G) and c.1827del (p.Tyr609X), confirming the diagnosis of LOPD.

Biological basis

Etiology and pathogenesis

Pompe disease is a hereditary condition transmitted as an autosomal recessive disease. The gene encoding acid alpha-glucosidase, GAA, has been localized on chromosome 17q25.3 (134). More information can be accessed at the following website:GAA Chromosome Locus.

Acid alpha-glucosidase is a lysosomal enzyme with both alpha-1,4-glucosidase and alpha-1,6-glucosidase activity and is, therefore, capable of digesting glycogen, maltose, and isomaltose completely to glucose (192). Like other lysosomal enzymes, acid alpha-glucosidase is a glycoprotein and is synthesized in the endoplasmic reticulum (ER)-Golgi complex. Before leaving the ER-Golgi complex to be transferred to the lysosomes, the inactive enzyme precursor of about 110 kDa is modified by proteolysis into a 95-kDa intermediate form, which is then further modified to the active forms of 76 and 70 kDa (166).

In agreement with the concept that there are no tissue-specific GAA isozymes, GAA activity in Pompe disease is markedly decreased in all tissues, not only in infantile Pompe disease but across the disease spectrum (58). The difference in clinical expression and pathology between infantile and late-onset forms of Pompe disease has been attributed to the presence of residual GAA activity in late-onset Pompe disease (LOPD) patients, but not in the infantile form. The difference in residual activity, first observed in muscle specimens, is also evident in cultured fibroblasts (230). Fibroblasts from patients with infantile Pompe disease have less than 1% residual activity, whereas fibroblasts from patients with late-onset disease have traditionally been thought to have between 1% and 40% residual activity. However, 13% of 101 LOPD had less than 1% residual activity in cultured skin fibroblasts in a study (12). Western blot analysis revealed that all of these cell lines produced GAA protein, but it was either exceedingly scarce or abnormally processed (12).

The improper processing, production, and activity of GAA across the Pompe disease spectrum is attributable to several factors, the most important of which is genotype. Pompe disease is a genetically heterogeneous disorder. Over 524 pathogenic variants in the GAA gene are listed on the Pompe Center website, which can be accessed at the following site:Erasmus MC: Pompe Center website. A study of 1079 patients from 26 countries reported 2075 GAA variants (80 novel) from the Pompe registry with the largest global repository data (201). Some variants are common in specific population groups, allowing correlations to be made between specific pathogenic variants and clinical severity (97). Other variants are rare, some being reported only in 1 family. The most common pathogenic variant in late-onset disease is c.-32-13T>G, which was first identified by Huie and colleagues and was found in 88% percent of the 101 patients with LOPD studied by Bali and colleagues (105; 12; 201). The c.-32-13T>G variant is the most commonly reported variant among patients with LOPD, globally and in specific geographic locations such as Europe, North America, and Latin America, but not in Asia-Pacific or Middle East (201). The c.-32-13T>G variant is called a “leaky splice site” variant because in 10% to 20% of splicing events, this “leaky” splice site variant allows some normal protein to be made, leading to its association with the milder, adult-onset disease (27). However, Herbert and colleagues characterized a cohort of 84 patients with the c.-32-13T>G variant who presented within the first 2 years of life with symptoms including delayed motor milestones, proximal weakness, swallow and feeding difficulties, and sleep apnea (91). This highlights early presentation of disease onset even in those with this pathogenic variant. Additionally, although cardiac involvement in the form of severe cardiomyopathy is rare in patients with LOPD with the c.-32-13T>G variant, arrhythmias have been reported in patients as young as 8 years old (235; 92). The c.525delT (9% U.S. cases, 34% Dutch cases) and exon 18 deletion (5% U.S. cases, 25% Dutch and Canadian cases) predominate among patients with IOPD (236; 97). Common variants due to founder effect in certain ethnicities include p.Asp645Glu in Chinese infantile patients, p.Arg854X in African Americans and Africans, and p.Gly309Arg in Europeans (121). The c.510C>T was found to be a genetic modifier in compound heterozygous and homozygous IVS1 patients (19). The c.510C>T reduces the extent of leaky wild-type splicing and results in early-onset of the disease, compared to other patients with LOPD (19).

Understanding of the correlation between the “molecular severity” of variants and severity of the clinical presentation continues to grow (67; 93; 133; 172; 131). This has been facilitated by documentation of clinical severity alongside specific variants in the Erasmus MC database, and by studies that transiently express pathogenic variant GAA in COS-7 and HEK293T cells and then analyze the activity, quantity, and quality of the enzyme (67; 93; 132; 133; 131). Although there is correlation between genotype and phenotype, other factors play a role in clinical presentation, emphasizing the clinical and genetic heterogeneity of Pompe disease and reinforcing the idea that disease presentation constitutes a continuum.

In a study of patients with infantile-onset Pompe disease (IOPD) by Bali and colleagues, certain GAA gene variants could be correlated to the patients’ cross-reactive immunological material (CRIM) status (11). CRIM status, an important factor in determining clinical response to ERT, is established by Western blot analysis: CRIM-positive indicates detectable GAA protein (unprocessed precursor band at 110 kDa or any of the processed forms) whereas CRIM-negative denotes absence of GAA protein (processed and unprocessed) (120). CRIM-negative patients typically mount an immune response (antibodies) against the infused rhGAA in ERT and have a poor clinical response to the subsequent doses of ERT. Approximately 25% of patients with classic infantile Pompe disease are CRIM-negative (11). The most commonly identified variants in CRIM-negative patients were p.Arg854X (32.7%) and c.525delT (4.8%). Forty-four out of 52 CRIM-negative patients tested homozygous or compound heterozygous for nonsense, frame shift, and multi-exon deletion variants consistent with the inability to make GAA protein. CRIM-positive patients, on the contrary, had 1 or 2 missense variants, in-frame deletions, or other variants that were expected to allow for some degree of GAA production. Berrier and colleagues showed that 16 out of 20 patients with IOPD (80%) who were CRIM-negative developed significant antibody titers, which led to invasive ventilation or death (21). There is growing evidence that a subset of patients with IOPD who are CRIM-positive also develop a significant immune response against the infused rhGAA. Desai and colleagues showed that about 12 out of 37 patients with IOPD (32%) who were CRIM-positive developed sustained antibody titers, which may result in a progressive clinical decline, despite being on enzyme replacement therapy (53).

In addition to findings of suboptimal levels of GAA activity, muscle biopsy shows a vacuolar myopathy across the Pompe disease spectrum. In the infantile form, many muscles contain an abundance of often-confluent vacuoles, resulting in a "lacework" appearance. There is accumulation of free and intra-lysosomal glycogen in most tissues that is especially severe in the heart. In agreement with the morphological appearance, glycogen content is massively increased in muscle from patients with IOPD, often reaching levels 10 times higher than normal. In LOPD, vacuoles are less numerous and tend to be smaller. The vacuoles contain periodic acid-Schiff-positive material and stain intensely for acid phosphatase, another lysosomal enzyme. Muscle biopsies from late-onset patients may appear normal, despite the marked decrease of GAA activity. The positive acid phosphatase stain can be a useful diagnostic clue in otherwise normal biopsy specimens. Muscle glycogen concentrations are generally lower in those presenting in childhood and may be normal in some patients presenting later in life. Thus, it is important to recognize that a normal muscle biopsy histology does not exclude a diagnosis of Pompe disease and, if it is suspected, further testing such as GAA enzyme activity analysis or gene sequencing is warranted.

As expected in a lysosomal disease, electron microscopy shows that much of the glycogen is contained within single membrane-limited lysosomal "sacs," but free glycogen is also present in the cytoplasm, especially in infantile-forms of GAA deficiency. The buildup of glycogen in the cytoplasm, which one might think would not occur due to the existence of cytoplasmic glycogenolytic enzymes, appears to be due to rupture/leakage of lysosomal glycogen (195). The lack of glycogen clearance from the cytoplasm can also be attributed to dysfunctional autophagy (171). Given that lysosomal storage of any substrate impairs autophagic delivery of bulk cytosolic contents to lysosomes (210), impairment of the autophagic pathway may help explain the relationship between different lysosomal enzyme deficiencies and cell death, secondary to autophagic buildup in all lysosomal disorders (209).

Impaired autophagy in Pompe disease is suggested by detailed studies both in the murine model (74) and in patients (152; 172; 171). In Pompe disease, buildup of autophagic debris in the skeletal muscle of patients and knockout mice appears to be correlated with a decrease in muscle strength, the accrual of defunct mitochondria, and mild muscular atrophy (195). In Pompe knockout mice, the loss of force calculated with a finite element model in isometrically contracting muscle slices containing different density and distribution of inclusions corresponded to the loss of force measured in ankle dorsal flexor muscles (60). Furthermore, complete removal of the genes for autophagy in mice resulted in death (126).

In a study examining human tissues, autophagic buildup was more prominent in adult fibers whereas the pathology of very young infantile patients (one as young as 1 month at baseline) primarily showed an abundance of over-expanded lysosomes at baseline (193). After 6 months of ERT, however, the infantile pathology started to reveal autophagic buildup as well. Different fibers tended to show either autophagic buildup or lysosomal expansion, suggesting a masking effect by the presence of glycogen. Importantly, a study highlighted the skeletal muscle pathology in patients with IOPD on long-term ERT (up to 8 years) and found that earlier treatment leads to less autophagic buildup (189). Another study reported the presence of an underlying pathology in 17 out of 23 patients (22 late-onset and 1 nonclassic infantile) (70). Muscle biopsies pre- and post-treatment with ERT revealed varying amounts of large, autofluorescent inclusions filled with lipofuscin, an indigestible material normally recognized in older patients as a result of aging. The ramifications of glycogen and autophagic buildup along with lipofuscin accumulation remain under investigation, both in relation to ERT efficacy and the prevention of normal contractile function.

Autopsy findings in long-term survivors with Pompe disease have helped to improve the understanding of the pathogenic process. Glycogen accumulation was found in smooth muscles of bladder and other parts of the genitourinary tract, which could account for the bowel dysfunction and urinary incontinence in patients with LOPD (182). In addition to the smooth muscles of bladder and smooth muscles of the gastrointestinal tract, glycogen accumulation in the respiratory system tissue, cardiac conduction tissue, iris sphincter, and blood vessels was observed. Skeletal muscles of tongue and upper esophagus were also found to have glycogen deposition.

Research on human-induced pluripotent stem cell-derived cardiomyocytes (iPSC-CM) found that Pompe disease iPSC-CMs produce lysosome-associated membrane proteins lacking glycosylation, revealing a Golgi-based glycosylation deficit most likely secondary to the changes in glycogen metabolism. These data suggest that Pompe disease cardiomyopathy has a glycan processing abnormality and thus shares features with hypertrophic cardiomyopathies observed in the congenital disorders of glycosylation (200).

Epidemiology

Prior to the advent of newborn screening, GAA deficiency was thought to have a combined estimated frequency of 1 in 40,000 live births and had been reported in many countries and in different ethnic groups (154). Certain pathogenic variants are more frequent in specific populations, often as a result of founder effect (67; 93). Using the 3 common pathogenic variants to screen an unselected sample of Dutch neonates, Ausems and colleagues calculated that the estimated incidence of infantile-onset Pompe disease (IOPD) was 1 in 138,000 and 1 in 57,000 for late-onset (09). In the United States, it was estimated that between 1900 and 3000 individuals were living with Pompe disease (154). By 2004, the estimated prevalence was between 5000 and 10,000 worldwide (247). Of the 4.1 million U.S. births recorded by the CDC in 2009, approximately 100 were expected to have Pompe disease, and about one quarter of those would be expected to have classic infantile Pompe disease.

The implementation of newborn screening (NBS) programs has helped to gather more accurate incidence data. A NBS program implemented in Taiwan uncovered that the prevalence of Pompe disease across the spectrum was approximately 1 in 18,108. Of those, 1 in 26,466 patients had late-onset Pompe disease (LOPD) and 1 in 57,343 had infantile-onset. As a result of NBS, the prevalence of pseudodeficiency also increased. Pseudodeficiency includes genetic variants that result in lower GAA activity, but no clinical manifestations. A high rate of pseudodeficiency has been found in Asian populations. The individuals with pseudodeficiency have a c.1726 G>A (p.G576S) variant in cis with c.2065 G>A (p.E689K), also known as the c.[1726A; 2065A] haplotype. Pseudodeficiency is common in the Asian population; it is seen in 3.9% of the healthy Japanese population and in 14.5% of the Taiwanese population (38). A report from newborn screening in Japan revealed low GAA activity in 71 of 103,204 (0.07%) newborns. Of these 71 patients, 32 (45.1%) were homozygous and 37 (52.1%) were heterozygous for pseudodeficiency alleles c.[1726G>A; 2965G>A] (164). Three newborns were identified with potential LOPD with no symptoms at the time of publication, and none were diagnosed with IOPD.

Since the addition of Pompe disease to the Recommended Uniform Screening Panel (RUSP), there is more information available about the incidence of Pompe disease from different states in the United States. Missouri was the first state to implement newborn screening for Pompe disease, followed by others. Implementation of NBS in the state of Missouri revealed an incidence rate of 1 in 5463 for Pompe disease. Eight patients were identified with a confirmed diagnosis of Pompe disease, of which 3 had infantile-onset, 3 had late-onset, and 2 had a variant of unknown significance/onset. Two patients were also identified with a pseudodeficiency allele, and 3 were carriers (102). A NBS program in the state of Washington screened approximately 110,000 newborns (208). The results of the identified data showed that 17 samples had low GAA enzyme activity, of which 4 had pathogenic variants consistent with Pompe disease, 4 were carriers, 3 carriers with a pseudodeficiency allele, and 6 samples were heterozygous for a pseudodeficiency allele only. The estimated prevalence of Pompe was calculated to be 1 in 27,800. In Illinois, after the implementation of newborn screening, 219,973 newborn dry-blood samples were screened by tandem mass spectrometry for lysosomal storage diseases, and an incidence of 1 in 21,979 was observed. Ten patients were identified with a confirmed diagnosis of Pompe disease, of which 2 infants had IOPD and 8 had LOPD. Fifteen patients had GAA pseudodeficiency, and 4 others had variants of unknown significance (30). Of 18,105 newborns screened in the New York pilot newborn screening program, 1 patient had a confirmed diagnosis of LOPD, 2 were carriers, and 3 carried the pseudodeficiency allele (244).

Prevention

Newborn screening helps in early detection of Pompe disease, thereby facilitating earlier and more effective treatment (214; 252). In the United States in March 2015 Pompe disease received approval from the U.S. Secretary of Health and Human Services to be added to the Recommended Universal Screening Program (RUSP) (144). At the time of publication, in addition to District of Columbia, 22 states have implemented universal newborn screening for Pompe disease, including Delaware, Washington, Oregon, California, Nebraska, Minnesota, Missouri, Illinois, Wisconsin, Michigan, Ohio, Kentucky, Mississippi, Tennessee, Virginia, Maryland, Pennsylvania, New York, Vermont, Massachusetts, Florida, and Rhode Island. Data from the newborn screening cohort in Taiwan continue to support that infants receiving enzyme replacement therapy prior to 1 month of age due to diagnosis via newborn screening show marked improvement (39). So far, data produced from the long-term follow-up of patients identified through this newborn screening program have demonstrated high and consistent efficacy of rhGAA treatment of classic IOPD. In 10 patients with left ventricular hypertrophy at diagnosis, all were able to walk independently after a median treatment time of 63 months, all had motor capability sufficient for participating in daily activities, and none required mechanical ventilation, but muscle weakness over the pelvic girdle appeared gradually after 2 years of age. Ptosis was also present in one half of the patients and speech disorders were common, suggesting some residual muscle weakness despite early initiation of ERT (40). In another study from Taiwan, newborn screening resulted in initiation of ERT as early as a median of 11.5 days. About 669,000 newborns were screened and 14 patients were diagnosed with Pompe disease. Physical examinations, blood tests, and echocardiography were performed within 2 hours of referral, and an early treatment with ERT was initiated (252). All 14 patients showed better outcomes in CK level, IgG antibody titers, left ventricular mass index, cognitive function (Bayley Scale of Infant and Toddler Development-3 scales) and motor function (Peabody Development Motor Scale, PDMS-2). Even among these 14 infants, they observed that starting ERT a few days earlier was associated with better biochemical responses and developmental outcomes.

As prenatal diagnosis is feasible using either biochemical and/or molecular genetic analysis, recurrence of infantile Pompe disease in the same family can be identified through antenatal detection and genetic counseling. There is a 25% recurrence rate with each pregnancy for couples identified as carriers due to the fact that GAA deficiency is inherited in an autosomal recessive manner (224). Of note, lifespan and phenotypic presentation among siblings with infantile Pompe disease is very similar (217). Siblings with LOPD, in contrast, can present very differently even though they share the same genotype. Phenotypic variation indicates the influence of epigenetic and environmental effects on clinical presentation and disease development (246).

Differential diagnosis

Despite these confirmatory methods of laboratory diagnosis, the rarity of Pompe disease and nonspecific presentation of many symptoms can complicate the clinical diagnostic process. The ACMG Practice Guidelines for Pompe disease offer a clear, concise method for diagnosis, both in infantile-onset and late-onset forms of the disease (121). The AANEM Consensus Treatment Recommendations for LOPD and Diagnostic Criteria for Late-onset (Childhood and Adult) Pompe Disease also provide useful methodologies (American Association of Neuromuscular & Electrodiagnostic Medicine 2009; 47).

Infants presenting with floppy baby syndrome should be referred for a chest x-ray. A finding of massive cardiomegaly is highly suggestive of classic infantile Pompe disease and separates Pompe disease from other causes of floppy baby syndrome, such as spinal muscular atrophy type I (Werdnig-Hoffmann disease) and other metabolic or congenital myopathies (104). An electrocardiogram (ECG) can further indicate the diagnosis, as infantile-onset Pompe disease (IOPD) patients typically have short PR intervals with tall QRS complexes. An echocardiogram would reveal a hypertrophic cardiomyopathy, occasionally with left ventricular outflow tract obstruction. Later in the disease course, IOPD patients may present with dilated cardiomyopathy and impaired cardiac function. Presymptomatic myopathy may be detectable via an electromyogram (EMG). Patients with IOPD will usually have high levels of creatine kinase (as high as 2000 UI/L) on examination (30). It should be noted, however, that this test is nonspecific as CK is typically elevated in other muscle diseases such as mitochondrial/respiratory chain disorders and several other glycogenoses and muscular dystrophies. AST, ALT and LDH may all appear elevated, and AST is typically higher than ALT. Urine Glc4 is a useful biomarker of overall disease burden and disease progression. In conditions with higher levels of glycogen storage or muscle breakdown, urine Glc4 levels increase correlatively (253).

Some infants with cytochrome c oxidase deficiency have both myopathy and cardiomyopathy, but cardiomegaly is usually less marked than in IOPD, and there is lactic acidosis. Deficiency of AMP-activated protein kinase, AMPK, or also known as PRKAG2 cardiac syndrome) can present with massive cardiomegaly in infancy, but muscle weakness is minimal and muscle biopsy is essentially normal (29). Another glycogenosis that can present with infantile muscle weakness and cardiomegaly is branching enzyme deficiency (glycogenosis type IV, brancher deficiency): the presence of abnormal glycogen (polyglucosan) in tissues of patients with branching enzyme deficiency distinguishes it from Pompe disease. Danon disease also resembles Pompe disease but is not due to GAA deficiency and is characterized by cardiomyopathy, intellectual disability, and autophagic vacuolar myopathy. This disorder is caused by pathogenic variants in the gene encoding a lysosome-associated membrane protein known as LAMP-2 (175).

In late-onset Pompe disease (LOPD), respiratory insufficiency may be the main complaint, accompanied by morning headache, exertional dyspnea, or sleep-disordered breathing (161; 02). A pulmonary function test revealing a decrease in FVC from the upright to supine position can suggest LOPD. Symptoms of muscles weakness typically trigger the initial diagnosis of either limb girdle muscular dystrophy (LGMD) or polymyositis. EMG features, especially in paraspinal muscles (fibrillation potentials, positive waves, complex repetitive discharges, and myotonic discharges), are useful clues to GAA deficiency (02). As patients with LOPD typically do not present with cardiac involvement, chest x-ray and electrocardiography were historically not been a part of the diagnostic process. However, with the emerging data showing Wolff-Parkinson-White syndrome, various other rhythm disturbances, and dilation of the ascending aorta as complications of LOPD, chest x-ray, echocardiogram, and electrocardiography have become important in the diagnostic work up and in the management of cases with LOPD (65). CK levels are usually elevated; however, about 5% of patients with LOPD have CK levels in the normal range (09). A study showed that diaphragm atrophy, often associated with reduced lung height and band-like atelectasis, can be considered a CT-MRI feature of respiratory insufficiency in LOPD (75). Early recognition of respiratory muscle involvement using CT-MRI could allow an early diagnosis and start of ERT in patients with LOPD. Literature also reveals the possibility of LOPD mimicking rigid spine syndrome in phenotypic presentation, as detectable using imaging data (177).

LOPD may simulate Duchenne muscular dystrophy in boys with calf pseudohypertrophy. In Duchenne muscular dystrophy, however, family history may suggest X-linked transmission, serum CK tends to be higher, and EMG does not show prominent irritative features or myotonic discharges. Muscle biopsy or blood-based enzyme analysis clearly differentiates the 2 conditions.

Other metabolic myopathies of childhood include debrancher deficiency (GSD IIIa), brancher deficiency (GSD IV), muscle phosphorylase b kinase deficiency, and carnitine deficiency. Debrancher deficiency myopathy is often accompanied by hepatomegaly, fasting hypoglycemia, and elevated cholesterol and triglycerides. There is no response of blood glucose to epinephrine or glucagon administration. Differential diagnosis from AMP-dependent protein kinase deficiency, brancher deficiency, and carnitine deficiency requires morphological and biochemical studies of muscle and enzymes.

As more information about the disease is emerging, it is apparent that late diagnosis and often misdiagnosis remains a problem in Pompe disease. It is important for clinicians to consider Pompe disease as a differential, when diagnosing myopathy or cardiac conditions. In a study involving 3076 adult patients who presented with hyperCKemia and/or limb-girdle muscle weakness, LOPD prevalence was found to be 2.4%. The most common clinical presentation included limb-girdle muscle weakness, respiratory insufficiency, and isolated hyperCKemia (150). A study showed that patients with Pompe disease who had a direct referral to expert clinics (metabolic or musculoskeletal) had the lowest diagnostic delay. In the study, 9 patients with IOPD and 23 with LOPD received a correct diagnosis at the expert centers; however, the patients who did not receive a direct and quick referral to the expert clinic had significant diagnostic delays (140). Similarly, lone cardiac involvement may be a preliminary finding in inherited metabolic disorders such as Pompe disease, and therefore, a cardiologist should have a high index of suspicion when the serum CK levels are higher than normal along with an abnormal echocardiography or rhythm disturbances, and subtle muscular issues (78).

Diagnostic workup

Confirmation of the diagnosis of Pompe disease requires demonstration of enzyme deficiency and/or presence of 2 pathogenic variants in trans in the GAA gene. There are several methods of measuring GAA enzyme activity, including assays on cultured skin fibroblasts, muscle biopsies, and blood (107; 186). Muscle biopsy, although invasive, provides material for histology and electron microscopy studies, in addition to GAA activity assay. This can be helpful, especially when the differential diagnosis is wide. Muscle biopsy can reveal glycogen accumulation in the lysosomes and in the cytoplasm (American Association of Neuromuscular & Electrodiagnostic Medicine 2009). However, it must be cautioned that, in late-onset Pompe disease (LOPD), different muscles tend to show different levels of glycogen accumulation, based on type of muscle fibers. In fact, histology from a certain area can be normal in a patient, and therefore, a normal histological study does not exclude a diagnosis of Pompe disease. Measurement of GAA activity in cultured skin fibroblasts is time consuming, requiring the time to culture the cells, but provides an accurate measurement of residual enzyme activity and material for cross-reactive immunologic material (CRIM)-testing. The cross-reactive immunologic material (CRIM) status has emerged as an important prognostic factor in patients with infantile Pompe disease and is now a part of the recommended diagnostic workup for babies with infantile Pompe disease identified via newborn screening (31). The presence of any amount of GAA activity on Western blot is considered as CRIM-positive, and a complete lack of the enzyme or the protein is considered as CRIM-negative. It is important to note that the CRIM status should not be interpreted as an “either/or” phenomenon, but rather a continuum. Hence, the CRIM status should be confirmed based on the Western blot and GAA variant analysis, or based on variant analysis alone if they are known pathogenic variants. The ability to accurately predict CRIM status based on underlying genotype is approximately 92%, and the turnaround time for this is very quick, often less than 48 hours. Research has presented a new blood-based CRIM assay that can yield results within 48 to 72 hours, providing potential opportunity for rapid classification of CRIM status to minimize treatment delay in patients with infantile-onset Pompe disease (IOPD) (242).

Variant analysis of the GAA gene allows identification of the associated DNA changes. GAA pathogenic variants may include missense, nonsense, splicing, frameshift, deletions, and insertions. GAA gene sequencing helps to confirm the diagnosis of Pompe disease, differentiate between IOPD, heterozygotes, and late-onset Pompe patients, and facilitate testing for family members, including carrier testing and prenatal testing (121). It also helps in determining CRIM status, which is critically important in the management of IOPD when initiating enzyme replacement therapy, and most recently, even gene therapy.

Genotyping also helps in identifying patients with pseudodeficiency of GAA activity, a complicating factor for newborn screening programs, as well as laboratory workup for some symptomatic patients. Pseudodeficiency of GAA has been linked to thep.Gly576Ser and p.Glu689Lys variants in cis, and it results in high false positives during NBS. To minimize these false positive tests, Taiwan adopted the inclusion of variant analysis in their newborn screening program to help differentiate pseudodeficiency from Pompe disease (137). An algorithm utilizing enzyme analysis, CRIM status analysis, and genotyping would be useful for NBS in order to minimize false positive rates and correctly identify pseudodeficiency in patients (15; 240). A study showed a second-tier of testing that could be added to the NBS and genotyping protocol to minimize the false positive rates from pseudodeficiency. This study included the use of Collaborative Laboratory Integrated Reports (CLIR), a new marker (ratio of creatine/creatinine (Cre/Crn) ratio as the numerator, and the activity of acid α-glucosidase (GAA) as the denominator), which successfully segregated Pompe disease and false-positive cases (229; 83).

Targeted gene sequencing allows sequencing of regions of particular interest and the identification of common or familial pathogenic variants. By focusing on the changes likely to be involved, resources are conserved. Targeted sequencing is important in situations where familial variants are known. However, a detailed family history should be collected from the proband and family to determine if more than 1 generation could be affected with Pompe disease. Careful consideration of family history or symptoms may warrant diagnostic evaluation for family members instead of targeted variant testing. There are reports of LOPD and IOPD in 1 family (100; 01; 138; 158).

Next-generation sequencing methods make it possible to test for several disorders simultaneously. In a study a gene panel comprising of 78 genes causing muscle disorders was used to analyze 34 patients with undiagnosed muscle disorders. The panel showed 100% sensitivity and 98% specificity across all genes. A diagnosis was reached in 32% of the patients. The panel achieved 100% sensitivity when used in 20 patients with Pompe disease with known variants (145). Another study successfully used next generation sequencing methods to differentiate patients with limb-girdle muscular dystrophies from patients with LOPD (both cohorts otherwise presenting with proximal muscle weakness) (23). This shows the potential value of next-generation sequencing in diagnosis of muscle disorders, especially in patients with nonspecific muscle weakness or atypical phenotypes.

Exome sequencing sequences all the expressed genes in a genome. This approach may be useful in undiagnosed neuromuscular disorders. The limitation of exome sequencing technology needs to be recognized. It does not reliably detect structural variants such as translocations or large deletions/duplications, and therefore, it can fail to detect certain types of pathogenic variants (225; 167).

Additionally, the diagnostic workup for Pompe disease should include determining the level of serum creatine kinase (31), which can range from normal to 15 times the upper limit of normal in Pompe disease (09). Serum creatine kinase is always increased in IOPD and can be normal in up to 5% of adult patients with LOPD. Therefore, creatine kinase should not serve as the only indicator of disease (121). Serum levels of creatine kinase, CK-MB, aspartate aminotransferase (113), alanine aminotransferase (197), lactate dehydrogenase, and urine Glc4 can also prove useful. AST and ALT are often elevated, with AST typically being the higher than ALT. Urinary Glc4 is a limit dextrin that is produced due to the breakdown of glycogen and is excreted in the urine. Urinary Glc4 is a good indicator of overall disease burden in Pompe disease because its level correlates well with the level of glycogen accumulation in skeletal muscles (253). Urinary Glc4 is found to be elevated as early as in neonates with IOPD. Therefore, it could be potentially used as a biomarker for IOPD at the time of NBS and to monitor response to enzyme replacement therapy (254; 40). Although the serum enzymes creatine kinase, CK-MB, AST, ALT, and LDH are useful in the diagnostic workup of Pompe disease, they are nonspecific biomarkers and should accompany other more specific diagnostic tests.

Electrocardiography and chest x-rays are useful for diagnosing IOPD. Electrocardiography shows short PR interval, giant QRS complexes, and signs of left ventricular or biventricular hypertrophy. In a study of 19 patients with IOPD, about 75% exhibited short PR intervals (06). Chest x-rays reveal cardiomegaly.

Pulmonary function testing should be done in both the supine and seated positions as it may reveal diaphragm weakness in LOPD (American Association of Neuromuscular and Electrodiagnostic Medicine 2009). Studies of pulmonary function show restrictive ventilatory insufficiency with reduced maximal static inspiratory and expiratory pressures, and early diaphragmatic fatigue. EMG of paraspinal muscles in LOPD may reveal "irritative" features and myotonic discharge. Magnetic resonance imaging of skeletal muscles may be useful to follow the progression of glycogen deposition and to monitor the response to therapeutic interventions (184; 59).

Whole-body magnetic resonance imaging (WBMRI) has also been used to measure the amount of fatty infiltration in skeletal muscles in LOPD. A study showed that a methodical evaluation of muscle strength and function (Gait, Stairs, Gowers, Chair (GSGC) score) and laboratory biomarkers (urinary Glc4) correlated well with the results from whole-body magnetic resonance imaging and, therefore, can be used to assess the severity of muscle disease and to follow patients with LOPD over time if unable to perform WBRI in patients with Pompe disease (117). Similarly, noninvasive 13C nuclear magnetic resonance spectroscopy can also help assess disease severity and follow progression and, if applicable, the effects of therapy (243). Glycemia and the response of blood glucose to epinephrine or glucagon administration are normal. Forearm ischemic exercise testing helps to exclude other glycogenolytic or glycolytic defects. These provocative tests are now not being done due to availability of noninvasive tests.

Management

Enzyme replacement therapy (ERT). Broad label FDA approval of ERT with alglucosidase alfa in 2006 and 2014 (Myozyme™ and Lumizyme™, respectively) changed the treatment landscape of Pompe disease. The 2 original pivotal studies on patients with infantile-onset Pompe disease (IOPD) uncovered that ERT improved survival, cardiac function, and both growth and motor development and that it was important to begin treatment as early as possible for maximum efficacy (119). For infantile Pompe disease, ERT has clearly shifted the natural history with survival now extending far beyond the first year of life: older patients in the first cohort are now as old as 19 to 20 years. Several long-term issues are emerging but, in general, the response of patients with infantile Pompe disease has been varied, as illustrated by several studies (122).

An important trend to note is that patients with IOPD typically reach a clinical plateau that requires an increase in ERT dose from 20 mg/kg every 2 weeks to 40 mg/kg every week, indicating the need for higher doses of ERT in children (189; 35). A study involving 8 children with classic infantile Pompe disease and CRIM positive status showed that starting at a higher dose of alglucosidase alfa (40 mg/kg/week) improved outcomes for ventilator-free survival and motor function compared to the current recommendation of 20 mg/kg every other week (241).

Another study of 7 children with IOPD and 4 with late-onset Pompe disease LOPD showed that there was a clinical plateau at standard dose of ERT and a higher dose of alglucosidase alfa up to 40 mg/kg every week was safe, which resulted in marked improvement in gross motor outcomes, lingual strength, pulmonary function measures, and biochemical markers (118).

The prolonged survival of patients with IOPD on ERT is providing new insight into the disease process as well as showing the true efficacy and limitations of ERT. Across the clinical spectrum, treatment response in patients continues to vary based on several factors, including muscle fiber type, defective autophagy, the degree of disease progression at time of treatment initiation, cross-reactive immunological material (CRIM) status, antibody response to ERT, underlying ACE variants, ACTN3 variants, gender, nutritional status, and insufficient enzyme uptake of skeletal muscle. Type II muscle fibers have been found to be less responsive to ERT than type I fibers (191; 194). This can explain why there is persistent weakness of certain skeletal muscles, such as the anterior tibialis in patients with IOPD who respond well to ERT (34). Case and colleagues postulate that weakness of the anterior tibialis, a muscle with a great number of type II fibers, may be attributed to the notable accumulation of autophagosomes and the subsequent accumulation of the recombinant enzyme within the autophagosomes preventing glycogen clearance in type II fibers. Female sex and younger age groups were found to be favorable prognostic factors for muscle strength and better clinical status for supine forced vital capacity respectively in a study by de Vries and colleagues (55).

Early treatment with ERT improves long-term outcomes in patients with Pompe disease, and the age at start of ERT is crucial to reduce muscular weakness and respiratory decline in IOPD (119; 36). Two children who started ERT late (7 and 8 months of age) became wheelchair-dependent and ventilator-dependent in their adolescence, and one child who started ERT at 3 months age was ambulatory at the age of 11 years (237). The United Kingdom experience with 20 infants treated between the years 2000 and 2009 was equally variable: overall ventilator-free survival was 35%, 35% died at a median age of 10 months, and 30% were alive but ventilator-dependent (36). Again, age and clinical severity at the beginning of treatment were of paramount importance. The results of the NBS program in Taiwan have conclusively shown that babies with IOPD who were identified through NBS and treated early did better than those identified clinically (250; 252). Fourteen infants with IOPD detected through the NBS program were treated with ERT as early as 11.92 (mean) days (range 6 to 23 days) of birth. All 14 infants had better biological, physical, and developmental outcomes, when compared to another group of infants with IOPD who began ERT 10 days later. Resolution of left ventricular hypertrophy, and normal motor and cognitive functions were observed in all 14 patients. Currently, the average age of ERT initiation is about 8 days of age (251). One noticeable benefit of earlier treatment in patients with IOPD, especially when compared to an untreated cohort, was the prevention of motor function deterioration (42). In a retrospective observational study of 14 patients with classic IOPD diagnosed via NBS, Chien and colleagues showed that there was a significant higher risk of motor decline in patients who had a delay in ERT-initiation or a delay to switch to higher doses of ERT (40 mg/kg) (43).

In addition to initiating care prior to disease onset, knowing CRIM status before starting ERT is critical in IOPD (120). Patients who are CRIM-negative tend to be part of the poor-responder subset, due to the development of high and sustained antibody titers (HSAT) to the alglucosidase alfa. One study showed that CRIM-negative infants with Pompe disease were much more prone to die or require invasive ventilation than CRIM-positive infants as a result of the immunological response to the exogenous protein (120). There is now emerging evidence that a subset of patients who are CRIM-positive (IOPD and LOPD) can also develop HSAT with a poor treatment outcome on ERT (13; 179).

Management of LOPD has its own challenges, and due to the slowly progressive nature of the disease, significant damage may already be present at the time ERT is started. ERT has proven effective in reducing disease burden in many cases of LOPD. A multicenter, randomized, placebo-controlled trial of alglucosidase alfa was conducted in 90 patients (8 years of age or older) with LOPD (239). The 2 primary end points were motor (6-minute walk test) and respiratory (percentage of predicted FVC) status. Secondary and tertiary end points were quantitative muscle testing and maximum inspiratory and expiratory pressures. The study concluded that treatment with ERT for 78 weeks was associated with improved walking distance and stabilization of pulmonary function over a period of 18 months. Similarly, an open-label, investigator-initiated, year-long observational study of ERT in 44 patients with LOPD led to the conclusion that the neuromuscular deficits were stabilized, and functional improvement was mild (222). Stabilization of disease progression and mild improvement of motor function was also noted by Bembi and colleagues (18). Another open-label study of 5 children (5.9 to 15.2 years) with Pompe disease showed stabilization of pulmonary function and increase of muscle strength (233).

Kuperus and colleagues compared clinical outcomes in 102 adult patients with LOPD treated with ERT to their expected untreated disease course extrapolated based on data collected prior to ERT initiation (135). Treated patients exhibited improved muscle strength (Medical Research Council sum score +6.6 percentage points, handheld dynamometry sum score +9.6 percentage points), improved pulmonary function (FVC upright +7.3 percentage points, supine +7.6 percentage points), and improved daily life activities (Rasch-Built Pompe-Specific Activity [R-PAcT] Scale +10.8 percentage points). Notably, the largest improvements compared to the natural disease course were seen during the first 2 to 3 years of treatment.

One study showed meta-data from 22 publications from 19 studies/trials that included 438 patients with LOPD (207). The study demonstrated a significant improvement in survival (5-fold improved compared to the untreated group). Pulmonary function declined consistently in the untreated group, which was in contrast to the ERT group, where patients with LOPD had an increase of 1.4% FVC as early as 2 months with ERT. Even though there was a slow regression toward baseline FVC values in 3 more years after ERT, the relative respiratory decline was significantly different from the untreated group. Ambulation and muscle strength also improved over time. The 6-minute walk test showed the highest level of improvement over the first 20 months of ERT with an increase by about 50 meters, which was followed by stabilization in the subsequent years on ERT. The untreated group did not show any improvement in the muscle strength testing (measured by the same 6-minute walk test) over time.

There was earlier identification of patients with LOPD via the NBS program (1 in 26,466 patients had LOPD) and earlier treatment in symptomatic patients. Four patients with LOPD were treated at ages 1.5, 14, 34, and 36 months, respectively, for symptoms including hypotonia, muscle weakness, elevated creatine kinase, and delayed developmental milestones. The data from this study showed that early treatment in patients with LOPD resulted in a good response to ERT (41).

Despite early initiation of ERT, there is growing evidence that survivors with IOPD and LOPD develop problems in the oropharyngeal areas, facial muscle weakness, articulation disorders, and sensorineural hearing loss (127; 251). Children with IOPD have early CNS white matter involvement as seen on brain magnetic resonance imaging, and they have relative weaknesses in certain domains of development and behavior (128; 129). Patients with IOPD and LOPD being treated with ERT have residual respiratory muscle weakness leading to much slower respiratory decline, when compared to the untreated group (207; 157; 251). Therefore, even though ERT is very beneficial for survival and reduces severe respiratory complications early in life, some extent of respiratory dysfunction exists, and this is a critical prognostic factor for Pompe disease (157).

Overall, it can be seen that patients with Pompe disease have benefited immensely from ERT, and that ERT has improved survival and quality of lives. For patients with LOPD, there are no consensus guidelines on when to begin ERT. The availability of NBS will help to understand this knowledge gap. However, to avoid irreversible damage, it is important to initiate ERT early in the disease course. In addition, due to the emerging phenotypes of IOPD survivors and ERT-treated patients with LOPD, it is clear that a more efficient therapy is the need of the hour. Multiple approaches have been explored either as adjunct therapies to ERT, second generation ERT, or gene therapy.

Adjunctive therapies. Diet and exercise play an important role in improving the clinical manifestations in patients with Pompe disease on ERT. The positive effects of a high-protein diet and aerobic exercise have been reported (213; 24; 17; 212). The validity of this therapeutic approach has been bolstered by a larger, though still uncontrolled, prospective study of 34 patients adhering to a high-protein and low-carbohydrate diet and exercise therapy (211), as well as a study of 44 patients with LOPD on ERT whose regular physiotherapy caused them to outperform their untrained peers in the 6-minute walk test (222). A 12-week exercise program in adult Pompe disease patients showed improvements in muscle pain and fatigue in addition to increase in aerobic fitness, muscle strength, and core stability (69). Aerobic exercise training combined with ERT has been shown to result in augmented glycogen clearance and removal of autophagic debris, and has a therapeutic potential for patients with Pompe disease (174; 228). Overexertion should not be part of the exercise regimen and program as it is thought that it can lead to increased muscle damage in those with disease-induced impairment of regeneration capabilities (73). The use of exercise as a therapy is called into question due to the potential deleterious effects of muscle contraction on lysosomes – stress on the lysosomes may cause rupture and glycogen leakage into the cytoplasm, furthering cell damage. On the other hand, pressure due to muscle contraction may help with glycogen clearance. Studies on the harm or benefit of muscular contraction in muscles with glycogen storage are needed in order to determine the optimal type and load of exercise necessary to build strength and, potentially, even help eliminate some of the glycogen buildup (34). Tarnopolsky and Nilsson have explained the theoretical basis for using mixed endurance and resistance exercises training, along with the benefits of protein and micronutrient supplementation in patients with Pompe disease. It is suggested that patients with LOPD should be encouraged to engage in low-to-moderate intensity exercises, under supervision on most days, and that that excessively strenuous regimen, excessive fatigue, and muscle pain should be avoided (228).

Respiratory muscle strength training is a form of pulmonary exercise that is being studied in individuals with Pompe disease. Patients with IOPD and LOPD on ERT have shown improvement in respiratory strength after undergoing respiratory muscle strength training (108; 110). Another study in patients with LOPD reported that improvements in inspiratory and expiratory muscle strength was sustained and durable after a 12-week respiratory muscle strength training program followed by 3-months detraining (109). There is an on-going, exploratory double-blinded, randomized control trial with an aim to provide calibrated, individualized, and progressive pressure-threshold resistance against inspiration and expiration using the respiratory muscle strength training program in patients with LOPD (111). In another study of 9 patients with LOPD on ERT, inspiratory muscle training for 8 weeks was found to increase the maximum inspiratory pressure from 3.3% to 50.5% (08).

Small molecule therapy. High doses of ERT are required to reduce the abnormal glycogen accumulation in skeletal muscles. This may be due to the low affinity of ERT for cation-independent mannose 6-phosphate receptor (CI-MPR) on the affected target cells (skeletal muscles) during endocytosis (257). To improve the delivery of ERT to the target cells, the addition of a high affinity, small molecule, chaperone therapy is being explored. Pharmacological chaperones include receptor agonists and antagonists, substrate analogues or other modulators, or ligands. Several preclinical studies have shown that beta blockers effectively improve the delivery of ERT into the skeletal muscles. An open-label pilot study of albuterol, when administered with ERT in 7 patients with LOPD who had previously been experiencing a clinical plateau on ERT, exhibited improvements in the 6-minute walk test, and several others improved in other parameters such as grip strength and the ability to stand up from a sitting position (123). Results from a 52-week, phase I/II study of clenbuterol in patients with LOPD treated with ERT showed improved 6-minute walk test distance, maximum inspiratory pressure, quick motor function test score, and gait, stairs, gower, chair (GSGC) test results (124). Clenbuterol decreased glycogen content in the vastus lateralis by 50% at week 52, and transcriptome analysis revealed more normal muscle gene expression for 38 of 44 genes related to Pompe disease following clenbuterol.

A phase 2a clinical trial demonstrated the use of a chaperone called Duvoglustat HCl (AT2220). It was coadministered as a 1-time single adjuvant drug along with the standard dose of ERT in 25 patients with Pompe disease. The results suggested a 1.2 to 2.8 fold increase in the GAA activity and protein, especially in the plasma and muscles in all 25 patients, without any adverse effects (122). Such improvements in the efficacy of traditional ERT therapies would benefit symptomatic patients and help in the prevention of long-term complications. In another study, the chaperone N-butyldeoxynojirimycin (D-NBDMJ) in combination with ERT resulted in a significant increase of alpha-glucosidase activity in 11 patients (178). The L-form (an enantiomer) of the same chemical compound, NBDMJ (ie, L-NBDMJ), was synthesized. This new enantiomer, L-NBDMJ, was shown to enhance lysosomal GAA levels in Pompe disease fibroblasts, either when administered in combination with ERT, or when given alone. In addition, it acted differently from its D-enantiomer and did inhibit glycosidase (48).

Immunomodulation with ERT. Immunomodulation is used to prevent or reduce the antibody response to ERT, thereby maximizing the effect of ERT. Attempts at immunomodulation in patients with Pompe disease on ERT result in the successful mitigation of negative responses. Immunomodulation with the anti-CD20 monoclonal antibody rituximab, methotrexate, and intravenous immunoglobulin has been successful in eliminating the deleterious antirecombinant enzyme antibodies in CRIM-negative patients (163; 15; 52). CRIM-negative status can often be predicted by the underlying GAA variants so that patients with the predisposition to develop antibodies to ERT can be identified and immunomodulation can be started simultaneously with ERT (11). It is extremely important to start immunomodulation prior to the formation of antibodies because the presence of high-sustained antibody titers diminishes ERT efficacy. This makes it hard and takes a longer time to abolish these entrenched immune response (13). Prophylactic, sustained immune tolerance induction (ITI) with rituximab, methotrexate, and intravenous immunoglobulin significantly reduced the left ventricular mass index, improved the overall survival, and eliminated/reduced antibodies against the ERT, making this immune tolerance induction protocol a relatively safe, feasible, and efficacious treatment modality to prevent immune responses to ERT among CRIM-negative patients with IOPD (114). Data from 34 ERT-naïve patients with IPD (25 CRIM-negative and 9 CRIM-positive) showed that a short-course prophylactic immune tolerance induction (with rituximab, methotrexate, and intravenous immunoglobulin) was beneficial, both in terms of safety and efficacy (52). Although CRIM-negative status is associated with a higher probability of antibody formation, a retrospective analysis by Banugaria and colleagues reported that 39% (9 out of 23) of CRIM-positive patients treated with ERT had high antibodies and poor clinical outcomes similar to those seen in CRIM-negative patients (14). In CRIM-positive patients, a prophylactic immune tolerance induction protocol with only transient, low-dose methotrexate has been safely administered, resulting in significant reduction in the number of patients developing high, sustained antibody titers and sustained intermediate titers when compared to a patient cohort treated with ERT monotherapy (115).

The use of bortezomib, a proteasome inhibitor, shows promise as an immunosuppressive agent for preexisting antibodies. The addition of bortezomib to immunomodulatory regimens resulted in a rapid reduction of HSAT (15). The reduction in antibody titers was also associated with an improvement in clinical outcome measures such as ventilator requirements, motor function, and cardiac parameters. A follow-up study demonstrated a durable and sustained immune tolerance to ERT in Pompe disease with entrenched immune responses using the bortezomib based immune tolerance induction protocol (116). Two of the 3 patients with IOPD were successfully weaned off all immune tolerance induction protocol medications (bortezomib, rituximab, methotrexate, and intravenous immunoglobulin) and continued to maintain low/no antibody titers. All 3 patients had a full recovery of their immune function. Another study highlighted the importance of a high dose of intravenous immunoglobulin as a lifesaving measure for a 7-year-old CRIM-negative IOPD patient who did not tolerate the immune tolerance induction protocol (196).

A number of other immunomodulation approaches have been reported in patients with Pompe disease including the use of rituximab, sirolimus or mycophenolate, and intravenous immunoglobulin (64; 51), plasma exchange, and omalizumab (203). The safety of rituximab in IOPD was described in 34 patients with IOPD (25 CRIM-negative and 9 CRIM-positive) for 11 to 55 weeks post-rituximab. The suppressed B-cells recovered in all children within a median time of 17 weeks (range 11 to 55 weeks), and 11 out of 12 children who were vaccinated with routine childhood immunizations, in the post-rituximab period, maintained humoral immunity (52). In February 2020, the Food and Drug Administration updated the prescription information for ERT to state that there is a risk of developing anti-rhGAA antibodies (IgG) against ERT in patients with Pompe disease. More information about Lumizyme can be accessed at the following website:http://products.sanofi.us/Lumizyme/lumizyme.pdf. The prescription information also specifies the need for routine monitoring of these antibody titers. An early assessment of the CRIM status in the disease process and the initiation of immune tolerance induction (ITI) protocol prior to and in conjunction with initiation of ERT are recommended. CRIM status is to be assessed for all patients every 3 months for the first 2 years (at the start of ERT), and then annually. Anti-rhGAA antibody testing may also be considered if patients develop hypersensitivity reactions, other immune-mediated reactions, or lose clinical response to ERT.

Second generation ERT. Studies are underway to enhance response to ERT through focusing on targeting and uptake of GAA via the addition of mannose 6-phosphate residues to the engineered alpha-glucosidase (256) and simultaneous administration of pharmacological chaperones to enhance delivery of ERT to the lysosomes (187).

NeoGAA. NeoGAA is a next-generation ERT wherein carbohydrates are conjugated to the GAA enzyme to improve its delivery to muscle cells by its increased affinity for cation-independent mannose-6-phosphate receptors (CI-MPR). A phase 1/2 study showed an improvement in the muscle strength and pulmonary functioning in patients who were ERT-naïve patients treated with neoGAA (183). A multicenter, international, phase 3 trial has been completed to evaluate the safety and efficacy of neoGAA infusions in patients with Pompe disease (NCT02782741). The preliminary data from this clinical trial suggest that NeoGAA is safe and well tolerated in patients with LOPD.

AT2221. An international, multicenter study is evaluating the safety of coadministration of a chaperone AT2221 (Miglustat) with a new recombinant human GAA in adult patients with LOPD (NCT04138277). The study showed that there were improvements in functional measures (motor function tests) and key biomarkers (CK, AST/ALT, urine Hex4) following 9 months of therapy. Among ERT-naïve patients, there was a 75 meter increase in 6MWT and 5% mean absolute change in percent predicted forced vital capacity, whereas among ERT-switch patients, there was a 37 meter increase in 6MWT and 2% mean absolute change in percent predicted forced vital capacity (04). At this time, a phase-3, open-label, international, multicenter study is evaluating the safety and efficacy of coadministration of the chaperone AT2221 with a new recombinant human GAA, in children with LOPD (12 to 18 years age) (NCT03911505).

Gene therapy. Gene therapy has shown promise to correct some of the chronic neuromuscular problems in Pompe disease that persist despite ERT, and it allows a long-term expression of a more potent therapeutic protein. Smith and colleagues conducted a human trial of diaphragmatic gene therapy (AAV1-CMV-GAA) to treat respiratory neural dysfunction in (ages 2 to 15 years). The subjects underwent a period of preoperative inspiratory muscle conditioning exercise (216). The change in respiratory function after exercise alone was compared to change in respiratory function after a combination of intramuscular delivery of AAV1-CMV-GAA to the diaphragm and continued exercise. At the end of 180 days, the maximal inspiratory pressure was unchanged in the combined intervention group; however, subjects in this group had improvements in flow and volume load compensation. The study found that subjects who benefited most tended to be younger, stronger, and used fewer hours of daily mechanical ventilation. The study concluded that combined AAV1-CMV-GAA and exercise training conferred benefits to dynamic motor function of the diaphragm, especially in children with a higher baseline neuromuscular function (216).

There are currently 4 ongoing clinical trials in phase I/II that are recruiting adults with LOPD for gene therapy. The trials can be accessed at the following websites: Clinical trial A, Clinical trial B, Clinical trial C, and Clinical trial D.

Stem cell therapy. Further advances in the field include the use of induced pluripotent stem cells (iPSCs) generated from the cells of patients with IOPD and LOPD. When treated with rhGAA, glycogen granules of the IOPD-specific iPSCs markedly decreased. Induced pluripotent stem cells can provide the opportunity to build up glycogen storage of Pompe disease in vitro (96).This offers promising insight into a future resource for the screening of new drug compounds and the development of new therapies for Pompe disease. Sato and colleagues generated LOPD-specific iPSCs and differentiated them into cardiomyocytes. A lentiviral vector, which expresses GAA, was used to infect the Pompe iPSCs. It was found that the lentiviral GAA transfer improved the disease-specific morphological changes such as glycogen accumulation and lysosomal enlargement (205).

Preclinical studies. The response of skeletal muscle to ERT is limited by the low expression of the cation-independent mannose-6-phosphate receptor (CI-MPR), a key molecule in the successful uptake of recombinant GAA enzyme (257). The role of CI-MPR was confirmed when the use of a selective beta-agonist, albuterol and clenbuterol, effectively increased skeletal muscle CI-MPR expression which, in turn, increased the efficacy of ERT in Pompe knockout mice (also known as GAA knockout mice) (141; 125). Clenbuterol, when administered with an AAV vector, was found to increase delivery of GAA to lysosomes. It was also found that clenbuterol reduced the glycogen content in skeletal muscle even in the absence of CI-MPR (68). In a study by Han and colleagues, salmeterol, another beta-agonist, was found to enhance the cardiac activity in GAA knockout mice in combination with an AAV vector encoding human GAA (141; 87). Propranolol, a commonly used nonspecific beta-blocker, was found to reduce glycogen clearance from skeletal muscle in GAA knockout mice, and reduce the efficacy of ERT, providing further evidence of the role of beta agonists (88). These studies have translated into clinical trials.

A study in GAA knockout mice showed a significant reduction in glycogen accumulation in the heart after the coadministration of DHEA, salmeterol, formoterol, or clenbuterol in combination with an adeno-associated virus (AAV) vector. Heart GAA activity significantly increased with the addition of salmeterol (87). Other approaches include substrate reduction therapy by inhibition of glycogen biosynthesis. Coadministration of mTORC1 inhibitor rapamycin with recombinant human GAA (rhGAA) decreased glycogen accumulation in muscles in Pompe knockout mice (07). Significant decrease in lysosomal glycogen accumulation in quadriceps, diaphragm, and heart was seen in Pompe mice models, which were administered rhGAA and an oligonucleotide suppressing muscle glycogen synthase (44). A study has identified therapeutic targets to reverse the aberrant mTOR signaling in Pompe disease by modulation of TSC2-Rheb pathway using Arginine (147).

It has been shown that nutrient-responsive transcription factors EB (TFEB) and E3 (TFE3) are recruited to lysosomes in nutrient-replete cells and play critical roles in nutrient sensing and regulation of energy metabolism. The overexpression of these transcription factors triggers lysosomal exocytosis and results in efficient cellular clearance of glycogen in Pompe cellular models. This offers TFEB and TFE3 as potential therapeutic targets for future advances in the management of Pompe disease (218; 153). At this time, no clinical trials of this concept have been undertaken.

To improve the specificity and efficacy of immune modulation in Pompe disease, a number of antigen-specific and antigen-targeted approaches have been investigated. For example, intravenous injection of synthetic vaccine particles carrying rapamycin during the first weeks of ERT in GAA knockout mice resulted in the inhibition of antibody formation. This effect was more durable than that of methotrexate treatment and resulted in improved glycogen clearance and motor function (146). Antigen encapsulation is also being investigated as a method to reduce antibody production to ERT. Pompe mice were intravenously injected with recombinant analogue alglucosidase-α encapsulated red blood cells that were artificially aged. The removal of aged red blood cells in spleen and liver allow contact with antigen presenting cells in these organs, allowing for immune tolerance to develop. The induced tolerance was found to be sustained for as long as 2 months (46). Oral administration of rhGAA also has been shown to reduce specific antibody formation against the enzyme in mice (176). Similarly, oral administration of GAA expressed in plant chloroplasts substantially suppressed GAA-specific IgG1 antibody formation in Pompe mice (223).

In preclinical studies, gene therapy has helped to improve the efficacy of ERT by improving enzyme delivery to target tissues. Gene therapy has also shown promise in preventing immune response to ERT with different approaches. In a preclinical trial, single administration of a non-depleting anti-CD4 monoclonal antibody (mAb) prior to administration of an AAV2/9 vector encoding GAA significantly reduced formation of anti-GAA IgGs in GAA knockout mice (86). In a follow-up study, a combination of combined anti-CD4 mAb and clenbuterol with an adeno-associated viral (AAV) vector, termed triple therapy, demonstrated a synergistic therapeutic efficacy on biochemical correction in cardiac and skeletal muscle. This triple therapy increased muscle strength as well as weight gain in GAA knockout mice (85). The efficacy of ERT against gene transfer with an AAV vector containing a liver-specific promoter was studied in GAA knockout mice (89). The study identified the minimum effective dose for effective AAV2/8-LPSPhGAA-mediated tolerogenic gene therapy in GAA knockout mice. The minimum effective dose was 8 x 1010 vg/kg to reduce glycogen content in the heart and diaphragm, and 3 times that dose was required to suppress previously formed anti-GAA antibodies to the ERT. Lim and colleagues demonstrated that a newly developed AAV-PHP.B vector was able to cross the blood brain barrier in GAA knockout mice and effectively reduce glycogen accumulation in the brain, heart, and skeletal muscles (148).

The quest to perfect the efficacy of ERT and develop new and improved therapies for Pompe disease continues. This is demonstrated with the ongoing clinical trials using different therapeutic strategies as well as preclinical work in the field of Pompe disease.

Special considerations

Adenotonsillectomy in children with IOPD. Children with IOPD have hypernasality and velopharyngeal incompetence that affects their speech and poses feeding difficulties (110). Generally, children with recurrent throat infections and sleep disordered breathing are often referred for adenotonsillectomy; one main complication of adenotonsillectomy is hypernasality. Therefore, children with IOPD who already have preoperative hypernasality can develop worsening of hypernasality postoperatively. Therefore, adenotonsillectomy should be avoided in children with IOPD, and other alternative approaches should be used to manage sleep disordered breathing (110).

Pregnancy

Pregnancy is a state of high physiological and metabolic demand. Women with LOPD have had successful pregnancies. Although there does not appear to be an increased risk of pregnancy or delivery complications associated with Pompe disease, there might be unmasking or worsening of symptoms of muscle weakness and respiratory complications during pregnancy in some women (112). There is limited experience regarding the effect of ERT on maternal and fetal outcomes.

In a study of 15 women with LOPD, a total of 36 pregnancies were reported in 13 women who were not on ERT. One subject conceived while on ERT and continued ERT through 2 normal pregnancies with worsening of weakness during pregnancy and improvement postpartum. In this cohort, there were 4 (11.1%) spontaneous first trimester miscarriages, 30 (83.3%) live births, and 2 (5.6%) stillbirths. The majority (75%) were vaginal deliveries and 6 (18%) were Cesarean sections. The preterm birth rate was 15.6% (5/32) and low birthweight rate was 12.5% (4/32). The data suggested that complications seen with pregnancy or childbirth were not higher in the study subjects, except for an increase in the rate of stillbirths. There was no difficulty carrying pregnancy to term (113).

There are published reports on 4 patients with LOPD who had successful pregnancy outcomes while on ERT. The course of all pregnancies was uneventful, fetal growth appeared normal in all gestational stages, and all delivered healthy babies. All children born to mothers with LOPD were healthy and continued to develop normally over time. Despite being on ERT, a decline in motor function and worsening respiratory function was reported, especially in the third trimester. Most patients regained their motor function, although not to prepregnancy levels in some. Respiratory function improved in the postpartum period in all, but did not always return to prepregnancy levels. Thus, a discontinuation of ERT during pregnancy is not recommended, as it can lead to progression of musculoskeletal and respiratory symptoms. In addition, it can make labor and delivery more challenging. In some pregnant women with LOPD who stopped ERT during pregnancy, emergence of allergic reactions on ERT reinitiation has been reported (202).

Experience from other lysosomal disorders where ERT is commonly continued during pregnancy, namely Gaucher disease and Fabry disease, is reassuring. Nevertheless, it must be recognized that dose of ERT in Pompe disease is up to 20-fold higher than that used in other lysosomal disorders. The Lumizyme package insert states that alglucosidase alfa should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus. However, based on our experience in Pompe disease and the risk benefit considerations, continuation of ERT appears to be very reasonable (101).

There are no particular dietary recommendations other than what is typically recommended during a pregnancy. As pregnancy is a state of high metabolic demand, a high protein diet should be followed.

Lactation. Alglucosidase alfa has been found to be secreted in breast milk, with activity levels in milk peaking 2.5 hours after the end of infusion. Activity returns to preinfusion levels 24 hours after the infusion. To minimize infant exposure to alglucosidase alfa, a nursing mother may temporarily pump and discard breast milk produced during the 24 hours after administration of alglucosidase alfa. There are no reports of adverse reactions due to increased enzyme activity in infants of these mothers and there is inadequate evidence to support complete abstinence from breast feeding (54).

Anesthesia

Patients with Pompe disease are known to be at increased risk during anesthesia. Heart conditions in pediatric patients with Pompe disease undergoing procedures should be monitored, as arrhythmias such as ventricular fibrillation have been documented (56). Malignant hyperthermia precautions should be taken. The use of propofol is not recommended as it may lead to myocardial ischemia through reducing myocardial contractility and systemic vascular resistance (106).

Organ donation. Very little information is available about the eligibility of patients with Pompe disease as organ donors. A donor with LOPD was successfully used for deceased donor liver and kidney transplantation. (84). Also, a 37-year-old, type 2 non-heart-beating donor with LOPD successfully donated both kidneys to 2 patients with chronic kidney disease (165). It is clear that with the increase of life expectancy in patients of Pompe disease further knowledge is required to understand the suitability of these patients as organ donors.