Early Detection of Pompe Disease by Newborn Screening Is Feasible: Results From the Taiwan Screening Program
OBJECTIVE. Pompe disease is an autosomal recessive lysosomal storage disorder that is caused by deficient acid α-glucosidase activity and results in progressive, debilitating, and often life-threatening symptoms involving the musculoskeletal, respiratory, and cardiac systems. Recently, enzyme replacement therapy with alglucosidase α has become possible, but the best outcomes in motor function have been achieved when treatment was initiated early. The aim of this study was to test the feasibility of screening newborns in Taiwan for Pompe disease by using a fluorometric enzymatic assay to determine acid α-glucosidase activity in dried blood spots.
METHODS. We conducted a large-scale newborn screening pilot program between October 2005 and March 2007. The screening involved measuring acid α-glucosidase activity in dried blood spots of ∼45% of newborns in Taiwan. The unscreened population was monitored as a control.
RESULTS. Of the 132 538 newborns screened, 1093 (0.82%) repeat dried blood-spot samples were requested and retested, and 121 (0.091%) newborns were recalled for additional evaluation. Pompe disease was confirmed in 4 newborns. This number was similar to the number of infants who received a diagnosis of Pompe disease in the control group (n = 3); however, newborn screening resulted in an earlier diagnosis of Pompe disease: patients were <1 month old compared with 3 to 6 months old in the control group.
CONCLUSIONS. To our knowledge, this is the first large-scale study to show that newborn screening for Pompe disease is feasible. Newborn screening allows for earlier diagnosis of Pompe disease and, thus, for assessment of the value of an earlier start of treatment.
- Pompe disease
- glycogen storage disorder type II
- acid α-glucosidase deficiency
- acid maltase deficiency
- enzyme assay
- newborn screening
- dried blood spots
Pompe disease, which is also called glycogen storage disorder type II and acid maltase deficiency, is a progressive, debilitating, and often fatal neuromuscular disease that is caused by deficient activity of the lysosomal enzyme acid α-glucosidase (GAA). In infants with Pompe disease, GAA activity levels in skin fibroblasts are typically <1% of the mean activity in normal control subjects, whereas in older children and adults, reported levels of activity range from 1% to 40% of that in normal control subjects.1–3 Pompe disease ranges from a rapidly progressive course, which is generally fatal by 1 to 2 years in infants, to a slower but nevertheless relentless, progressive course that results in significant morbidity in adults. In infants and young children, Pompe disease is characterized by prominent hypotonia, muscle weakness, motor delay, feeding problems, and respiratory and cardiac insufficiency.1,2 A retrospective study found that the median age at symptom onset was 2 months, the median age at diagnosis was 4.7 months, and the median age at death was 8.7 months.2 Enzyme replacement therapy with recombinant human GAA can be used to treat patients with Pompe disease and has been shown to prolong survival, reverse cardiomyopathy, and improve motor function.4,5 The best motor function outcomes have been achieved when enzyme replacement therapy was initiated early, which underscores the need for early diagnosis4,6,7; however, early diagnosis of infantile-onset Pompe disease is usually not possible because of the low index of suspicion, lack of specificity in its early symptoms, and use of traditional GAA activity assays that require growth of fibroblasts from skin biopsies.
GAA deficiency results in progressive lysosomal glycogen accumulation primarily in muscle cells. Muscle biopsies from 8 patients (2.7–14.8 months old) showed that 25% to 58% of the total tissue area contained glycogen.8 In infants with Pompe disease, enzyme replacement therapy was most successful in those who were treated earlier; after 52 weeks of enzyme replacement therapy, repeat biopsies showed a marked reduction of glycogen content.8 Until recently, demonstration of deficient GAA activity in dried blood spots (DBSs) was not possible because of interference from the isoenzyme maltase glucoamylase (MGA), which is abundant in neutrophils. The identification of maltose and acarbose as effective inhibitors of MGA permits the assessment of GAA activity in blood samples, including DBSs on filter paper.3,9–11 These methods have been shown reliably to identify patients with Pompe disease in small cohorts of infants9,11 and adults3 and to provide a rapid way to diagnose Pompe disease.
This is the first report of a large pilot program to screen for Pompe disease in newborns by using a fluorometric enzymatic assay to determine GAA activity in DBSs on filter paper. The objective of this study was to determine whether this technique is effective in screening for Pompe disease in a newborn screening program.
The newborn screening pilot program started in October 2005 and was conducted by the newborn screening center of the National Taiwan University Hospital (NTUH), which screens ∼45% of all newborns in Taiwan. Samples are typically collected within 3 days of birth. Because NTUH is also the referral center for clinical diagnosis of Pompe disease, the unscreened population was monitored for comparison. This article reports the results from October 2005 to March 2007 (18 months), but the program is ongoing. Informed consent was obtained for each sample collected and assay performed. The DBSs used in this study were the ones collected for routine newborn screening.
Three assays were performed: (1) GAA activity, measured at pH 3.8 in the presence of acarbose; (2) total GAA (tGAA), measured at pH 3.8 without acarbose; and (3) total neutral glucosidase activity (NAG), measured at pH 7.0 without acarbose. The tGAA reflects the combined activity of the isoenzyme MGA and GAA and was measured to calculate the percentage of tGAA that was inhibited by acarbose by using the formula (tGAA − GAA)/tGAA. The NAG was measured to control for the quality of the sample and to calculate the ratio of NAG to GAA.
For establishment of a normal population mean, GAA activity was measured in 5000 anonymous newborn samples. Another 2000 anonymous newborn samples were used to obtain a new population mean each time the protocol was modified.
For screening, a 2-tiered method was used (Fig 1). Samples with GAA activity <55% of the normal mean in the first tier were retested in the second tier for GAA and NAG activity. When the second tier screen showed GAA activity <25% of the normal mean and an NAG/GAA ratio >25, a second DBS was obtained and tested for GAA, tGAA, and NAG; when the NAG/GAA ratio was >100, the newborn was recalled immediately for confirmatory testing. When GAA activity was <8% of the normal mean, percentage of tGAA inhibition was >80%, and the NAG/GAA ratio was >60 in the second DBS, the newborn was brought in for confirmatory testing. Finally, GAA deficiency was confirmed when GAA activity in mononuclear blood cells was <5% of the normal mean.
GAA Activity Assays
For determination of GAA activity, a modified protocol that was based on the fluorometric assays developed by Chamoles et al9 and Kallwass et al3 was used. Briefly, one 3.2-mm-diameter disk from each DBS specimen was extracted with deionized water in black 96-well assay plates (Corning 3650, [Corning, Corning, NY]) for 1 minute at room temperature followed by gentle mixing on a rocking platform at 4°C for 1 hour.
A 70-mM stock solution of the synthetic substrate 4-methylumbelliferyl-α-D-glucoside (Calbiochem, San Diego, CA) in dimethyl sulfoxide (EM Science, San Diego, CA) was prepared in advance. Substrate solutions at pH 3.8 and pH 7.0 were prepared by a 50-fold dilution of this stock solution with 40 mM aqueous sodium acetate buffer (pH 3.8) and 40 mM sodium acetate buffer (pH 7.0), respectively.
Enzyme reactions at pH 3.8 and pH 7.0 were composed of 50 μL of substrate solution, 10 μL of deionized water, and 40 μL of DBS extract. For enzyme reactions in the presence of inhibitor, the water was replaced by 10 μL of aqueous acarbose (40 μM; Toronto Research Chemicals, Toronto, Ontario, Canada). These reagents were incubated for 20 hours at 37°C covered with sealing film (Corning 6570). The DBS extract for blanks was incubated separately and combined with the other reagents at the end of the incubation period, immediately followed by addition of 200 μL of 150 mM EDTA (pH 11.5) to all wells. A 4-methylumbelliferone standard curve was prepared on every plate by mixing 100 μL per well aqueous standards in the range of 0.00 to 3.13 μM with 200 μL per well EDTA solution. Eight different standards per curve were used in duplicate. In addition, low controls (from patients with Pompe disease) and high controls (from healthy infants) were added to each assay plate.
Molar product quantities in the assay wells were calculated by linear regression from the standard curve; GAA activity is presented as μmol/L whole blood (Wb) per hour. For this, it was assumed that a 3.2-mm punch contained 3.0 μL of Wb. Filter paper discs from a single venous blood specimen were included as internal quality control samples on all plates. The quality control was monitored closely, and all changes in the assay protocols were noted: in November 2005, the concentration of acarbose was reduced from 40 μM to 4 μM; in June 2006, the protocol was changed from manual operation to automated operation by Freedom EVO100 (Tecan, Durham, NC); and in February 2007, a Biomek NXP (Beckman Coulter, Fullerton, CA) was used.
When a newborn was recalled for confirmation, a 7-mL blood sample was drawn: 2 mL was used immediately to perform blood chemistry assays, including creatine kinase and creatine kinase myocardial band. The remaining 5 mL was used to obtain purified lymphocytes to perform a GAA activity assay. A physical examination looking for cardiac murmurs, congestive heart failure, hypotonia, muscle weakness, and crying cyanosis was performed, and an electrocardiogram and chest radiograph were taken immediately. On any suspicious finding, an echocardiogram was conducted immediately; enzyme replacement therapy with alglucosidase α (Myozyme [Genzyme, Cambridge, MA]) was started within 1 week of the enzymatic confirmation of diagnosis for patients with confirmed cardiac involvement. Genetic counseling and family history were also discussed as part of the confirmatory process. A baseline study, which included muscle biopsy and skin biopsy, was conducted before the first infusion with alglucosidase α.
GAA Activity Assay in DBSs
The within-run precision was 0.14% (coefficient of variation [CV]: 15 runs with n = 16 in each run). The day-to-day precision was 8.72% (CV: 15 runs), and the operator variability was 2.06% (CV: 8 for 1 manual operator and 7 for 1 automated liquid handler).
The distribution of GAA activities in the original DBSs of the screened population was similar in each month (Fig 2). Except for the change of acarbose concentration in November 2005, changes in the protocol have little impact on changes in the mean GAA activity of the screened newborn population and of the normal population mean (Fig 2).
The GAA activity assay results showed a normal distribution (Fig 3A); the majority of samples tested (84.4% of the screened population) had GAA activity levels within 55% to 200% of the normal mean (Fig 3B). Of the screened population, 13.3% and 2.3% had a GAA activity of <55% and >200% of the normal mean, respectively. For these newborns, a second-tier test on the original DBS was performed.
Results were reported for 132 538 newborns; 8 (0.006%) were recalled immediately for confirmatory testing because their NAG/GAA ratio was >100. A second DBS from 1093 (0.82%) newborns was tested, and positive results were found for 113 (0.085%) newborns. Confirmatory testing revealed that of the 121 newborns (8 of whom were recalled immediately after the first screening test), 117 were presumed to have been identified by false-positive results: mean GAA activity in the first DBS of these newborns was 1.45 μmol/L Wb per hour (range: 0.18–3.24 μmol/L Wb per hour), which is lower than the current normal activity of 16.52 μmol/L Wb per hour (range: 1.78–40.76 μmol/L Wb per hour) but overlaps with the current range of the first DBS for Pompe disease (0.38–2.34 μmol/L Wb per hour). No false-negative results were observed in this pilot study.
Confirmation of Pompe Disease
Four newborns had confirmed GAA deficiency (newborn screened and diagnosis of Pompe disease confirmed 1 [NBS1], NBS2, NBS3, and NBS4); NBS2, NBS3, and NBS4 had an NAG/GAA ratio >100 and were referred directly after the first screening. Table 1 gives an overview of the results of the GAA activity assays in DBSs, lymphocytes, and skin fibroblasts. Newborn NBS1 had no clinical symptoms, (ie, normal cardiac function and normal muscle strength) when a diagnosis of Pompe disease was confirmed at 40 days of age; therefore, enzyme replacement therapy was not initiated. By the age of 9 months, NBS1 had developed clinical symptoms of axial muscle weakness, and treatment was started at 14 months of age.
A diagnosis of Pompe disease was made at 19, 22, and 9 days for NBS2, NBS3, and NBS4, respectively. There were no obvious clinical symptoms, but abnormal signs included cardiomegaly, ventricular hypertrophy, an elevated left ventricular mass index, and increased creatine kinase levels (Table 2). Although muscle strength was normal for all cases diagnosed within the newborn screening program, muscle biopsies showed vacuoles in 15% to 85% of the myocytes in NBS2, NBS3, and NBS4 (Table 2). Figure 4 shows the confirmatory chest radiograph and muscle biopsy of NBS3 to illustrate the cardiomegaly and affected myocytes. Directly after confirmation of the diagnosis of Pompe disease, NBS2, NBS3, and NBS4 were started on treatment with enzyme replacement therapy. Results of this will be discussed in a subsequent article.
The GAA assays in DBSs, lymphocytes, and fibroblasts were also used to confirm the diagnosis of Pompe disease in 3 newborns from the control group (CLIN1, CLIN2, and CLIN3) who were referred to the NTUH (Table 1). All 3 were clinically normal at birth but later developed symptoms suggestive of Pompe disease, except for CLIN3, who had a heart murmur. Compared with the affected newborns who were identified by screening, these newborns started enzyme replacement therapy at a later age (4.2, 5.8, and 2.9 months, respectively) and had more severe symptoms (fever, cough, poor appetite, cardiomegaly, left ventricular hypertrophy, failure to thrive, and hypotonia in newborn CLIN1; and bronchiolitis, cardiomegaly, congestive heart failure, and hypotonia in newborn CLIN2; Table 2). Their newborn screening DBSs were obtained after consent from their parents. Data from the analyses of these DBSs indicated that these cases could have been detected by screening at the time of birth (Table 1).
This is the first large, population-based study to show that newborn screening for Pompe disease is feasible. With the use of the 2-tiered screening algorithm, a similar number of newborns were identified compared with the number of patients who had Pompe disease and were found in the rest of Taiwan, but the age at which their diagnosis of Pompe disease was confirmed ranged from 9 to 22 days, compared with 3 to 6 months for the control group. This earlier diagnosis allows for early enzyme replacement therapy with alglucosidase α, which can be life-saving and could prevent irreversible muscle damage.4,6,7 Currently, the results from the first-tier assay are reported within 7 days after the arrival of the samples; however, the physiologic high level of creatine kinase in newborns who are <2 weeks of age will interfere with the confirmation of diagnosis of Pompe disease at this early age.
This project also increased our understanding about the early manifestations of Pompe disease in young infants. In the past, early symptoms of Pompe disease were often derived from recall of symptoms from affected patients or from occasional early detection of affected siblings. In this study, we demonstrated that cardiac involvement was detectable in all 3 cases before the age of 1 month. Moreover, although none of the screened newborns who received a diagnosis of Pompe disease had weakness of skeletal muscles, biopsies already revealed prominent muscle involvement; therefore, the results of this study clearly indicate that the pathogenesis of Pompe disease in both cardiac and skeletal muscles occurs long before the appearance of clinical symptoms.
In this study, 117 samples in which GAA activity was lower than the current reference range were identified, but these infants did not show signs of hypotonia or cardiomyopathy; therefore, these 117 samples were presumed to reflect false-positive results. Because it is yet unknown whether a subset of newborns who are identified as having false-positive results will eventually develop Pompe disease at a later age, especially 3 infants in whom a low GAA activity in skin fibroblasts has been confirmed, these infants will be closely monitored for symptoms of Pompe disease. Furthermore, the probability of these infants' carrying pathologic mutations or polymorphisms of the GAA gene is being investigated. Identification of later onset Pompe disease in newborns may be stressful to the patients and their families; however, the ability to monitor the patients and to intervene by alleviating symptoms through enzyme replacement therapy before irreversible muscle damage occurs can reduce this stress.
The modified GAA assay that was used in this study is sufficiently robust for newborn screening. The decrease of acarbose concentration from 40 μM to 4 μM resulted in a higher average GAA activity but allowed for a better discrimination between patients and control subjects. Lysosomal enzymes seem to remain stable in DBSs for sufficiently long periods, as has been previously reported.9,13,14 In our study, storage of DBSs up to 4 weeks at −20°C did not affect the GAA activity. The GAA assay can be performed by using commercially available reagents and may be readily incorporated into current newborn screening programs. The 2-tier design of the test significantly decreases the load of the screening laboratory, because the second-tier test, which involves the simultaneous measurements of 2 enzymes, is required for only a small proportion of the newborns. The recall rate of the screening for Pompe disease is not significantly different from other screening tests,15 including screening tests for cystic fibrosis (0.6%)16 or congenital adrenal hyperplasia (0.74%).17 If the target of screening is limited to infantile-onset Pompe disease, not including the milder variants, then the recall rate may be further decreased. In Fig 5, we present data from an 8-month period when both the NAG/GAA ratio and tGAA inhibition were measured. We also plot data from the newborn DBSs for the 7 patients detected during the whole study period (NBS1–4 and CLIN1–3). We can see that the lowest NAG/GAA ratio among the patients is 40; therefore, it is safe to elevate the cutoff for the NAG/GAA ratio from 25 to 30, which would be expected to decrease the recall from 0.83% to 0.37%.
In this study population, the prevalence of Pompe disease was 1 in 33 333 (95% confidence interval: 1:12 048 to 1:100 000). In general, the prevalence of Pompe disease is reported to be between 1 in 40 000 and 1 in 300 000, depending on geographic and ethnic factors.1 Although a study conducted in northern Portugal reported a prevalence of 1 in 600 000,18 other studies, conducted in the Netherlands and New York, found frequencies of 1 in 40 000.19,20 Similar to this, the frequency of Pompe disease in newborns in southern China and Taiwan is estimated to be 1 in 20 000 to 1 in 40 000.21 More prolonged studies will be necessary to give a more precise prevalence of Pompe disease in Taiwan. Furthermore, the advent of improved screening methods may help to provide a more accurate picture of the prevalence of Pompe disease in different regions. Together with the recently available enzyme replacement therapy, this may ultimately prompt a reassessment of current screening policies.
Other methods have been developed to screen newborns for Pompe disease by using fluorometric immune quantification13 or tandem mass spectrometry.10,14 Both methods allowed for GAA assessment as a component of a multiplex assay for multiple lysosomal storage disorders, which is advantageous for large-scale screening purposes. The tandem mass spectrometry method measures GAA activity, whereas the immune-quantification method measures the amount of enzyme present, rather than enzyme activity. The latter method could lead to false-negative results for newborns who produce normal levels of defective enzyme and false-positive results for newborns who produce reduced levels of enzyme with normal activity. Additional research is needed to determine whether 1 of these techniques is preferable for assaying GAA activity alone or in combination with other enzyme assays.
This study in a large population shows that newborn screening for Pompe disease is feasible. The GAA assay in DBSs is a fast and consistent method and may allow for the identification of infants with Pompe disease before clinical symptoms manifest. Given that early diagnosis of Pompe disease can lead to earlier treatment initiation, which is an important predictor of treatment efficacy, it seems that newborn screening for Pompe disease is warranted.
This study was supported by an unrestricted research grant from the Genzyme Corporation (Cambridge, MA) and by the NTUH A1 project.
- Accepted January 17, 2008.
- Address correspondence to Wuh-Liang Hwu, MD, PhD, National Taiwan University Hospital, Department of Pediatrics, 7 Chung-Shan South Rd, Taipei 10016, Taiwan. E-mail:
Financial Disclosure: Drs Chen and Hwu have received research/grant support from Genzyme Corporation for other research or activities not reported in this article; Dr Y.T. Chen has served as a consultant for Genzyme Corporation; and Drs Keutzer and Zhang are employees of Genzyme Corporation. The other authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Pompe disease is a progressive and often fatal lysosomal storage disorder. Two methods to detect GAA (the deficient enzyme) in dried blood samples have been developed, and studies with small numbers (<1000) of infants or adults have shown promising results.
What This Study Adds
This is the first large-scale study to show that newborn screening for Pompe disease is feasible. Newborn screening allows for earlier diagnosis of Pompe disease and, thus, assessment of the value of an earlier start of treatment.
- ↵Hirschhorn R, Reuser AJ. Glycogen storage disease type II: acid alpha-glucosidase (acid maltase) deficiency. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease. 8th ed. New York, NY: McGraw-Hill; 2001:3389–3420
- ↵Kishnani PS, Corzo D, Nicolino M, et al. Recombinant human acid [alpha]-glucosidase: major clinical benefits in infantile-onset Pompe disease. Neurology.2007;68 (2):99– 109
- ↵Van den Hout JM, Kamphoven JH, Winkel LP, et al. Long-term intravenous treatment of Pompe disease with recombinant human alpha-glucosidase from milk. Pediatrics.2004;113 (5). Available at: www.pediatrics.org/cgi/content/full/113/5/e448
- ↵Li Y, Scott CR, Chamoles NA, et al. Direct multiplex assay of lysosomal enzymes in dried blood spots for newborn screening. Clin Chem.2004;50 (10):1785– 1796
- ↵Lin CY, Shieh JJ. Molecular study on the infantile form of Pompe disease in Chinese in Taiwan. Zhongua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi.1996;37 (2):115– 121
- Copyright © 2008 by the American Academy of Pediatrics