The effect of enzyme replacement therapy on clinical outcomes in paediatric patients with Fabry disease – A systematic literature review by a European panel of experts

Background: Fabry disease is caused by a deficiency of the lysosomal enzyme α-galactosidase, resulting in progressive accumulation of globotriaosylceramide (GL-3). The disease can manifest early during childhood and adolescence. Enzyme replacement therapy (ERT) with recombinant human α-galactosidase is the first specific treatment for Fabry disease and has been available in Europe since 2001. This paper presents the findings of a systematic literature review of clinical outcomes with ERT in paediatric patients with Fabry disease. Methods: A comprehensive systematic review of published literature on ERT in Fabry disease was conducted in January 2017. The literature analysis included all original articles reporting outcomes of ERT in paediatric patients. Results: Treatment-related outcomes in the paediatric population were reported in six publications derived from open-label clinical trials and in 10 publications derived from observational or registry-based studies. ERT was shown to significantly reduce plasma and urine GL-3 levels in paediatric patients with Fabry disease. The effect of ERT on GL-3 clearance from renal podocytes appeared to be agalsidase dose-dependent. ERT relieved pain and improved gastrointestinal symptoms and quality of life. Conclusions: Based on the published literature, the use of ERT in paediatric patients can significantly clear GL-3 accumulation, ameliorate the early symptoms of Fabry disease, and improve quality of life. Treatment with ERT in paediatric patients with Fabry disease may be important to prevent further disease progression and overt organ damage. List of abbreviations BPI Brief Pain Inventory BRIEF Behaviour Rating Inventory of Executive Function CHQ-P50 Child Health Questionnaire Parent Form 50 ECG electrocardiogram eGFR estimated glomerular filtration rate EOW every other week ERT enzyme replacement therapy FOS Fabry Outcome Survey GFR glomerular filtration rate GI gastrointestinal GL-3 globotriaosylceramide HRV heart rate variability HUI2/3 Health Utility Index 2/3 https://doi.org/10.1016/j.ymgme.2018.04.007 Received 5 April 2018; Accepted 20 April 2018 ⁎ Corresponding author at: Department of Paediatrics, University of Torino, Piazza Polonia 94, Torino 10126, Italy. E-mail address: marco.spada@unito.it (M. Spada). Molecular Genetics and Metabolism 126 (2019) 212–223 Available online 26 April 2018 1096-7192/ © 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). T LNR laboratory normal range LVH left ventricular hypertrophy LVM left ventricular mass LVMi left ventricular mass index lyso-GL-3 globotriaosylsphingosine mGFR measured glomerular filtration rate PedsQLTMCFS PedsQLTM Cognitive Functioning Scale pNN50 percentage of successive normal sinus RR intervals> 50ms QoL quality of life QSART quantitative sudomotor axon reflex testing QTc heart rate-corrected QT interval r-MSSD root mean square of the successive differences SDNN standard deviation of the normal-to-normal interval VAS visual analogue scale


Introduction
Fabry disease (OMIM #301500), an X-linked lysosomal storage disorder, is caused by mutations in the GLA gene (OMIM #300644; HGNC 4296) encoding the lysosomal enzyme α-galactosidase resulting in the accumulation of glycosphingolipids, such as globotriaosylceramide (GL-3) and globotriaosylsphingosine (lyso-GL-3), in a wide range of cell types [1]. The accumulation of glycosphingolipids triggers a cascade of events causing cellular dysfunction and progressive damage to multiple vital organ systems including the heart, nervous system, and kidneys [1]. Individuals with mutations associated with the classic phenotype generally experience symptoms from early childhood onwards. The earliest symptoms seen in affected paediatric patients typically include neuropathic pain, hypohidrosis, heat intolerance, and gastrointestinal (GI) complaints such as nausea, abdominal cramps, and diarrhoea. In adults, there are manifestations of progressive disease in multiple organ systems including albuminuria/proteinuria, decline in glomerular filtration rate (GFR), end-stage kidney disease, left ventricular hypertrophy (LVH), cardiac arrhythmia, heart failure, transient ischaemic attack, and stroke [1][2][3][4][5][6][7][8][9][10][11].
Two forms of human α-galactosidase enzyme replacement therapy (ERT) are available for the treatment of Fabry disease: agalsidase alfa (Replagal®), administered at a dose of 0.2 mg/kg every other week (EOW); and agalsidase beta (Fabrazyme®) administered at a dose of 1.0 mg/kg EOW. Agalsidase alfa and agalsidase beta are both available in most European and Asian countries, and in Australia and Canada, but only agalsidase beta is approved by the US Food and Drug Administration (FDA) (in 2003) [12].
Previous literature reviews reporting on the efficacy and safety of ERT in Fabry disease have analysed data derived from randomized controlled clinical trials only [13][14][15]. To our knowledge, there is no comprehensive report covering all types of original publications, and the whole range of treatment outcomes across all organ systems.
In 2017, a comprehensive systematic literature review of all published original articles on ERT in the treatment of Fabry disease was conducted. This paper summarizes the findings of the literature review with respect to treatment outcomes in paediatric patients.

Methods
A full description of the methodology used in the systematic literature search and analysis also appears in a related article in this issue [16]. The findings of the literature analysis for male [17] and female [18] populations, as well as a position statement on therapeutic goals in Fabry disease based on the outcomes of an expert consensus panel [19] also appear in this issue. This article reviews publications retrieved in the literature search that describe ERT outcomes in the paediatric population (patients ≤ 18 years of age) published up to 31 January 2017.
The outcomes that were selected for analysis included plasma and urine GL-3 and lyso-GL-3, tissue histology, measures of kidney and cardiac function, and cardiac morphology. Other outcomes included nervous system parameters as well as pain, GI outcomes, and quality of life (QoL). Regarding GL-3 values, if GL-3 was reduced by ERT treatment, the values are reported as described in the publication. Normalization is only reported if it is described as such in the publication, due to differences in assays and cut-off levels for GL-3.
Results are described for the approved dose regimens: agalsidase alfa 0.2 mg/kg EOW and agalsidase beta 1.0 mg/kg EOW, unless otherwise specified. Publications describing the results of studies in which data from patients treated with agalsidase alfa and agalsidase beta were combined, or in which the ERT type was not specified, are referred to in the analysis as "mixed-ERT" publications.

Results
A total of 34 publications included in the systematic literature analysis reported ERT outcomes data in paediatric patients (Table 1). Publications were available from clinical trials (CT), observational studies (OS), and case reports (CR). Studies describing both treatments (i.e. mixed-ERT studies), or when treatment was not specified (i.e. ERT NS), were also included.
For agalsidase alfa, there were 5 CT publications (all of which were Grade 1c single-arm clinical trials), four OS publications (all of which were Grade 3 retrospective observational studies), and 6 CR publications (two grade 4 case series, and four Grade 5 case reports). Two CT publications [22,23] were extension trials of study [21]. Patients lost to follow-up in the first extension study were due to study site closures and differences in treatment duration due to staggered study entries [22]. The second extension study only included patients who had participated in two separate phases of the study, prior to (phase 1) and after (phase 2) manufacturing changes in agalsidase alfa [23].
For agalsidase beta, 1 CT publication of a Grade 1c single-arm clinical trial was included, as well as 2 OS publications (both Grade 3 retrospective observational studies), and 8 CR publications (one Grade 4 case series, seven Grade 5 case reports).
No clinical trial studies were performed with mixed or unspecified ERT regimens; 4 OS publications were included (two Grade 2 prospective observational studies and two Grade 3 retrospective observational studies), as well as 4 CR publications (two Grade 4 case series and two Grade 5 case reports).
The main findings regarding clinical outcomes of treatment with approved doses of agalsidase beta or agalsidase alfa in paediatric patients with Fabry disease are summarized in Table 2. 3.1. GL-3 accumulation 3.1.1. Plasma GL-3 3.1.1.1. Agalsidase alfa 0.2 mg/kg EOW. Treatment with agalsidase alfa was associated with plasma GL-3 reductions in five single-arm CT publications (Table 3) [20][21][22][23][24]. Of these, one CT followed 24 paediatric patients for 6 months [21], of which 17 were then followed for 48 months [22] and 11 followed for 78 months [23] in extension trials. The first publication showed that agalsidase alfa was associated with significant reductions after 6 months in plasma GL-3 levels, which were elevated at baseline in males [21]; these reductions were sustained for up to 48 months [22]. In the 78-month follow-up report of this CT [23], plasma GL-3 levels did not decrease further. In contrast, no change was observed in the females, the majority of whom did not have abnormally elevated GL-3 levels at baseline [21,22]. In a report of a single-arm CT involving 13 paediatric patients [20], male patients had slightly elevated GL-3 levels at baseline, which were reduced after 5.3 months of treatment, whereas female patients did not have elevated plasma GL-3 levels at baseline. In another single-arm CT publication including 14 paediatric patients treated for 12.6 months, plasma GL-3 Table 1 Publications reporting clinical outcomes of ERT in paediatric patients with Fabry disease. The low-dose group received (1) agalsidase alfa 0.2 mg/kg EOW (n = 5) or (2) agalsidase beta 0.2 mg/kg EOW (n = 1) throughout the study. The high-dose group received (1) agalsidase beta 1 mg/kg EOW (n = 3) for the entire study period or (2) agalsidase beta 1 mg/kg EOW after 4 years of agalsidase alfa 0.4 mg/kg EOW (n = 1), agalsidase alfa 0.4 mg/kg EOW (n = 1), or agalsidase alfa 0.2 mg/kg, on a weekly basis (n = 1) for the entire study period. d Case 1 received 5 years of agalsidase beta 1.0 mg/kg EOW, 3 years of agalsidase alfa 0.2 mg/kg EOW, and 2 years of agalsidase beta 1.0 mg/kg EOW. Case 2 was treated with 5 years of agalsidase beta 1.0 mg/kg EOW, had a 7-month treatment hiatus, received agalsidase alfa 0.2 mg/kg EOW for 2 years and 5 months, and then was switched to agalsidase beta 1.0 mg/kg EOW. The regimen for Case 3 was as follows: 5 years of agalsidase beta 1.0 mg/kg EOW, 13 months of agalsidase beta 0.5 mg/kg EOW, 2 years of agalsidase alfa 0.2 mg/kg EOW and was then switched back to agalsidase beta 1.0 mg/kg EOW. levels were elevated at baseline in both male and female patients. A significant decrease was observed in male patients only [24]. One CR supported reduction of elevated plasma GL-3 at baseline levels with ERT [33].
3.1.1.2. Agalsidase beta 1.0 mg/kg EOW. For agalsidase beta, one singlearm CT publication reported plasma GL-3 levels in paediatric males and females (Table 3). In males, the levels were elevated at baseline, normalized by week 4, and remained normal thereafter. Levels in females were normal at baseline and remained unchanged during   [22] of the clinical trial [21]. The baseline data for plasma GL-3 in [22] however are presented for the total group of 24 patients from the original RCT [21] of which 19 were male. Case series, case reports, mixed-ERT publications, paediatric-adultmixed publications, and publications with other dose regimens are not included. Results are summarized as increase (↑), decrease (↓), or no change from baseline to follow-up after ERT initiation. Data are means (SD) or means ± SE or medians (range), unless otherwise indicated.
treatment [35]. An OS publication of agalsidase beta in 10 paediatric patients showed that elevated plasma GL-3 levels decreased towards the normal range in male patients, and normalized in female patients after 2-33 months of treatment (Table 3) [36]. Three CRs also reported normalization of plasma GL-3 levels with agalsidase beta [38,42,44].

Mixed ERT.
Two OS publications showed that elevated plasma GL-3 levels were reduced in 9 or 10 paediatric patients after 12-60 months of treatment with ERT [46,47].  Table 1). Of these, two single-arm CT extension trials of study [21] showed significant reductions from baseline levels at 18 months in males [22], and were sustained for up to 78 months [23]. Two other CT publications reported higher baseline urinary GL-3 levels in males than in females; decreases from baseline were observed in both male and female patients [20,24]. One CR showed a decline in urinary GL-3 [33].

3.1.2.2.
Agalsidase beta 1.0 mg/kg EOW. One OS publication on agalsidase beta reported that, in the three paediatric males with elevated urinary GL-3:sphingomyelin ratios at baseline, levels declined towards normal values after 5-27 months of treatment. In one female, a slightly elevated urinary GL-3:creatinine ratio at baseline normalized during follow-up (Supplementary Table 1) [36]. One CR also described normalization of urinary GL-3 levels [38].

Mixed ERT.
Two OS publications showed that elevated urinary GL-3 levels were reduced in 9 or 10 paediatric patients after 12-60 months of treatment with mixed ERT [46,47].

Urinary lyso-GL-3.
No publications reporting on the change in urinary lyso-GL-3 levels in paediatric patients following agalsidase treatment were found.  ) showed that all patients had GL-3 inclusions in renal cells. Seven patients did not receive ERT prior to renal biopsy, whereas two patients (both male) received agalsidase alfa for 2 years prior to renal biopsy, and were the only patients with no GL-3 inclusions in endothelial cells of the kidney [29]. Additionally, another CR examining pre-treatment renal GL-3 accumulation and pathology in eight paediatric patients reported data from one patient who had renal GL-3 levels re-assessed after initiation of ERT. This patient had an unchanged, high podocyte GL-3 accumulation score after 3 years of agalsidase alfa [34].
3.1.3.2.3. Mixed ERT. One OS of 12 young classic Fabry patients (nine of whom were ≤18 years of age) investigated kidney GL-3 accumulation during ERT with either agalsidase alfa 0.2 or 0.4 mg/kg EOW or agalsidase beta 0.2 or 1.0 mg/kg EOW [47]. After 65 months of treatment, complete clearance of GL-3 from glomerular epithelial and mesangial cells was reported in all patients. In podocytes, the clearance of GL-3 was ERT dose-dependent: patients who received agalsidase 0.2 mg/kg EOW did not achieve significant GL-3 clearance from podocytes, whereas patients receiving agalsidase beta 1.0 mg/kg EOW showed significant clearance [47].
In a CR, composite podocyte scores decreased in three patients as compared with baseline after 5 years of agalsidase beta 1.0 mg/kg EOW, and increased again after switching to 3 years of agalsidase alfa 0.2 mg/kg EOW [50]. The GL-3 clearance from podocytes was most notable in the patient who started ERT at a younger age (7 years), compared with the patient who started ERT later (18 years) [50].
3.2.1.2. Agalsidase beta 1.0 mg/kg EOW. One 11-month single-arm CT publication of 16 paediatric patients [35], one OS of 10 patients with 12-96-month data [36] and one CR (3 years of treatment [43]) reported no change in eGFR values, which were normal at baseline. One OS publication reported normal eGFR at baseline, which remained normal after over 60 months of agalsidase beta treatment (Supplementary Table 3) [37]. [47]. Similarly, another OS study found that mixed-ERT treatment did not result in significant eGFR changes [48]. One CR reported that in one male patient treated with agalsidase beta 1.0 mg/kg EOW for five years, mGFR values decreased on switching to agalsidase alfa 0.2 mg/kg EOW [50].

Albuminuria
3.2.2.1. Agalsidase alfa 0.2 mg/kg EOW. One single-arm CT publication reported normal albuminuria following 6 months agalsidase alfa treatment in the 16 of 24 patients who had normal albuminuria at baseline [21]. The single-arm CT extension phase of this study reported a slight increase in albuminuria levels up to 48 months of treatment in 10 paediatric patients [22]. Another single-arm CT publication in which all patients had microalbuminuria at baseline reported no changes in microalbuminuria after 12.6 months of ERT [24]. An OS FOS registry publication reported no change in the number of patients with microalbuminuria during 24 months of treatment (12 patients at baseline, 11 patients at 24 months) (Supplementary Table 4) [27].

Mixed ERT.
The impact of ERT on the albumin:creatinine ratio might be dose-related. An OS reported significantly decreased albumin:creatinine ratios with agalsidase beta (1.0 mg/kg EOW), but not with agalsidase alfa (0.2 mg/kg EOW) [47]. In one CR, the normal albumin excretion levels at baseline generally remained stable during ERT [50].
3.2.3. Proteinuria 3.2.3.1. Agalsidase alfa 0.2 mg/kg EOW. Two single-arm CT publications reported normal proteinuria levels in patients without proteinuria at baseline [20,21]. An extension CT publication reported that none of the eleven paediatric patients who were normoproteinuric at baseline developed proteinuria at the last follow-up at 78 months [23]. Another single-arm CT publication in which all patients had microalbuminuria at baseline reported no changes in proteinuria levels after 12.6 months of treatment [24]. In a FOS registry study, proteinuria was observed in one patient at baseline and after 12 months; three other patients also developed proteinuria after 24 months of treatment (Supplementary Table 4) [27]. Two case reports noted stable proteinuria levels in the normal range during agalsidase alfa treatment [31,33].

3.2.3.2.
Agalsidase beta 1.0 mg/kg EOW. One single-arm CT publication including 16 paediatric patients (14 males) reported a slight reduction in mean urinary protein excretion after 11 months of treatment with agalsidase beta [35]. Mild proteinuria (> 100 mg/m 2 /24 h) was present in 8 of 15 evaluable patients before treatment, but in only three patients at the end of follow-up [35]. An OS publication reported that proteinuria remained within the normal range in 4 males during 65-126 months of follow-up (Supplementary Table 4) [37]. One CR stated no change in proteinuria [39], but another showed temporary increases in proteinuria when the agalsidase beta dose was reduced from 1.0 mg/kg EOW to 0.7 mg/kg once a month or 0.6 mg/kg EOW [44].

Mixed ERT.
One CR described a 16-year old boy with severe proteinuria which decreased to normal levels after 1 year of ERT and adjuvant therapy [51]. Another CR of an 11-year old girl with early signs of renal involvement as demonstrated by the presence of proteinuria described that proteinuria levels improved after starting ERT (no follow-up specified) [53].

Cardiac outcomes
3.3.1. Left ventricular mass (LVM) and other cardiac echocardiography and magnetic resonance imaging outcomes 3.3.1.1. Agalsidase alfa 0.2 mg/kg EOW. Three clinical trial publications of a paediatric population followed for 6, 48, and 78 months reported normal LVM or LVM index (LVMi) at baseline and stable LVM/LVMi during ERT [21][22][23], an effect that was also observed in another singlearm CT after 12.6 months of treatment [24]. An OS FOS registry publication reported that five of the six paediatric patients with echocardiography assessments had LVMi above the 75th percentile for their age. Mean LVMi decreased during treatment [26]. Another OS FOS registry publication reported normal LVMi in 40 children at baseline, which remained normal and stable in the 17 children followed for 24 months (Supplementary Table 5) [27]. One OS described three males who developed LVH during treatment (Supplementary Table 5) [28]. LVM measurements were stable and normal in one CR [33].
3.3.2. Other cardiac outcomes 3.3.2.1. Agalsidase alfa 0.2 mg/kg EOW. One single-arm CT publication reported no change in mid-wall fractional shortening after 12.6 months of agalsidase alfa [24]. Effects of ERT on ventricular wall thickness were not described in this population and left ventricular ejection fraction data were reported in only one single-arm CT publication, where normal ejection fraction value at baseline and during 6 months of ERT was observed [21].

Electrocardiogram (ECG) measures
3.3.3.1. Agalsidase alfa 0.2 mg/kg EOW. Two single-arm CT publications including a 6 month study and a 48 month extension study of paediatric patients reported no overall change from normal baseline values in ECG parameters of heart rate, PR interval, QRS duration, and heart rate-corrected QT interval (QTc) (Supplementary Table 5) [21,22]. This effect was also observed in another single-arm CT publication [20]. One OS publication including 22 treated and 44 untreated patients reported the appearance of frequent ventricular premature beats in two treated patients during the third year of agalsidase alfa treatment and supraventricular premature beats with T-wave inversion in one treated patient during the second year of treatment [28].
One single-arm CT publication described that at baseline, heart rate variability (HRV) was generally lower in males compared with females. After 6 months of agalsidase alfa treatment, HRV increased significantly in males and remained normal in females [21]. An extension study including 11 patients showed that HRV improvements remained stable after 78 months of treatment [23]. One single-arm CT publication reported no changes in HRV, which was normal in most cases at baseline, measured using three different parameters after 12.6 months of agalsidase alfa treatment (Supplementary Table 5) [24].

3.3.3.2.
Agalsidase beta 1.0 mg/kg EOW. Published data from two single-arm CT studies of agalsidase beta reported normal ECG findings at baseline and no significant overall changes in ECG parameters after up to 96 months of treatment (Supplementary  Table 5) [35,36].

Mixed ERT.
One OS publication reported the occurrence of significant bradyarrhythmia that required implantation of a pacemaker in one patient despite 5.5 years of ERT, comprising 26 months of agalsidase alfa 0.2 mg/kg EOW and 41 months of agalsidase alfa 0.4 mg/kg EOW. After implantation, the patient received agalsidase beta 1.0 mg/kg EOW [47].
3.4. Nervous system outcomes 3.4.1. Sweat function 3.4.1.1. Agalsidase alfa 0.2 mg/kg EOW. One single-arm CT publication reported an increase in sweat volume after 6-12 months of agalsidase alfa as measured by quantitative sudomotor axon reflex testing (QSART) or with a Macroduct® sweat collector [20]. This effect was also observed in two CT publications examining paediatric patients after 6 months [21] and after 6 and 12 months [22] of treatment. The single-arm extension CT publication, however, reported decreased sweat volume after 3.5 years of treatment. There were only six paediatric patients with data available at this time-point (Supplementary Table 6) [22]. Two CR publications report treatment with agalsidase alfa also described improved sweating in paediatric patients with ERT [32,33].
3.4.1.2. Agalsidase beta 1.0 mg/kg EOW. One OS examining treatment with agalsidase beta in 10 paediatric patients reported increased sweat function in the majority of male patients after 1 year of treatment, which was sustained throughout the treatment period (up to 96 months) [36]. In two female patients, sweat function decreased (Supplementary Table 6) [36]. One CR reported normalization of sweating after 1 year of agalsidase beta [45].  Table 6) [27]. One CR described a rapid improvement in heat intolerance in a male patient after initiation of treatment [30], and another noted that the ability to sense heat and cold improved during treatment in one male patient [32]. 3.4.2.3. Mixed ERT. One CR described heat intolerance in a female patient, which improved with ERT treatment [53].
An OS FOS registry publication reported the hearing function at baseline and after 24 months of treatment in 39 paediatric patients: hearing impairment was reported in 7 patients at baseline and in 4 patients after 24 months of treatment, and tinnitus in 14 patients at baseline and in 11 patients after 24 months of treatment (Supplemental Table 6) [27]. 3.4.4.2. Mixed ERT. One OS publication reported significantly higher cognitive functioning in paediatric patients treated with ERT compared with those who did not receive treatment with ERT. Cognitive function was assessed using the Behaviour Rating Inventory of Executive Function (BRIEF) and PedsQL™ Cognitive Function Scale (PedsQL™ CFS), but, as a limitation of the study, the treated group was not assessed at baseline [49].

Pain outcomes
3.5.1. Agalsidase alfa 0.2 mg/kg EOW Pain outcomes assessed using Brief Pain Inventory (BPI) scores were reported in two single-arm CT studies [20,24], as well as two CT singlearm studies following a paediatric population for 48 months after an initial six month trial [21,22]. One of these did not report significant decreases in BPI "pain at its worst" scores in both male and female paediatric patients after 6 months of treatment, but in the same study 6 out of 11 patients were able to reduce or stop the use of neuropathic pain medications [21]. The publication that reported 12-month pain outcomes in the same cohort did show significant decreases in both "average pain" and "pain at its worst" BPI scores in males [22]. One single-arm CT publication (13 patients) reported a reduction in the number of patients with neuropathic pain after 5.4 months of treatment [20] and an OS FOS registry publication reported a reduced incidence of pain crises in female paediatric patients during 24 months of treatment [27]. This publication also reported chronic pain in nine patients at baseline and in four patients after 24 months of treatment (Supplementary Table 7) [27]. The effects on pain from CRs were variable, with one showing a decrease in pain following treatment [30], while another reported a dose reduction of neuropathic pain medication with carbamazepine (from 750 to 300 mg daily) following ERT [32], and another an increase in pain [33].

Agalsidase beta 1.0 mg/kg EOW
One OS publication assessed pain outcomes using visual analogue scale (VAS) scoring in patients treated with agalsidase beta 1.0 mg/kg EOW [36]. All patients reported acroparaesthesia at baseline. After 12-96 months of ERT, neuropathic pain was ameliorated as indicated by reduced VAS scores from 6.5 and 6.6 to 1.0 and 2.5 in male and female paediatric patients, respectively (Supplementary Table 7) [36]. The effects of agalsidase beta on pain outcomes were variable in CRs: one reported an increase in pain [44] whereas most described a decrease in pain [38,40,42,43,45] and one reported a reduction of neuropathic pain medication with phenytoin (from 200 mg/day to 100 mg/ day) and carbamazepine (from 400 mg/day to 200 mg/day) after enzyme treatment [41].

Mixed ERT
A five-year follow-up OS publication showed a substantial reduction in Fabry disease severity score of the peripheral nervous system (DS3-PNS score) of which pain is an integral factor, in eight of the nine ERTtreated children studied [47]. Two CRs noted an improvement in pain outcomes [52,53]. Another CR reported that acral pain decreased during 5 years of treatment with agalsidase beta 1.0 mg/kg EOW, and increased again with treatment discontinuation, agalsidase beta dose reduction (0.5 mg/kg EOW), or when patients were switched to agalsidase alfa 0.2 mg/kg EOW [50].
3.6. Gastrointestinal (GI) outcomes 3.6.1. Agalsidase alfa 0.2 mg/kg EOW A cohort of 13 patients reported variable effects of treatment on GI outcomes in one single-arm CT study [20]. Two OS FOS registry publications reported improvements in abdominal pain and diarrhoea in children treated for up to 24 months of treatment [25,27]: one of these publications, which included 127 paediatric patients (14 with available GI data at 12 months, 10 with available data at 24 months), showed a decline in GI pain from 80% to 50% of patients after 24 months of treatment (Supplementary Table 8) [25]. A CR described improvements in GI events [30].

Agalsidase beta 1.0 mg/kg EOW
A single-arm CT of 16 paediatric patients treated with agalsidase beta reported a significant improvement in postprandial pain and vomiting, and a non-significant decline in nausea [35]. In another OS publication, VAS scores for abdominal pain were reduced in both male and female patients following 12-96 months of treatment [36] (Supplementary Table 8) [36]. In one CR, diarrhoea resolved after 1 year of agalsidase beta treatment [45].

Mixed ERT
One CR noted improvements in GI events after ERT initiation [53]. In another CR, GI outcomes (abdominal pain, diarrhoea) decreased with 5 years of agalsidase beta (1.0 mg/kg EOW), but increased when treatment was stopped, the dose was reduced, or the regimen was changed to agalsidase alfa (0.2 mg/kg EOW) [50].

mg/kg EOW
In a single-arm CT study including 24 paediatric patients, QoL was assessed using the Health Utility Index (HUI) 2 or 3 questionnaires [21]. Near-normal QoL scores were observed at baseline and during 6 months of treatment [21]. In another single-arm CT study, there was little change in QoL scores (HUI2/HUI3 and Childhood Health Questionnaire physical and psychosocial scores) after 12.6 months of treatment (Supplementary Table 9) [24]. An OS FOS registry publication (98 paediatric patients) described fatigue in 14 patients at baseline and in seven patients after 24 months of treatment [27].

Agalsidase beta 1.0 mg/kg EOW
In a single-arm CT study of agalsidase beta, 48 months of treatment in 16 paediatric patients (14 males) was associated with improvement in a range of QoL parameters including significantly reduced proportion of days absent from school due to illness, increased proportion of days where patients could report good general health, and a decrease in the number of days when the patients experienced difficulty in performing low, medium, and high-energy activities [35]. An increase in energy levels and the ability to perform physical exercise in 10 paediatric patients (six males) was also reported in another OS publication (Supplementary Table 9) [36].

Mixed ERT
No publications describing the change in QoL following treatment were identified.

Discussion
This systematic literature review investigates the effect of ERT on clinical outcomes in paediatric patients with Fabry disease as reported in six publications derived from clinical trials, and in 10 publications derived from observational studies. ERT significantly reduced or normalized plasma GL-3 levels, relieved pain, improved gastrointestinal symptoms, and increased quality of life [20][21][22]25,27,35,36,47]. Some patients were able to reduce or discontinue the use of pain medication [21,32,40,41,50]. The management of pain is particularly important in children [36] as pain reliefor decrease in use of symptomatic neuropathic pain-control medications -can reduce sedation, improve concentration, and lessen school absences [35].
In one publication, the youngest patient to receive agalsidase beta treatment had almost complete clearance of podocyte GL-3, suggesting a potential relationship between age of ERT initiation and extent of podocyte GL-3 clearance [47]. Considering that the podocyte's integrity is important to prevent proteinuria, the timely initiation of treatmentbefore the onset of irreversible pathological changesmay contribute to the long-term positive prognosis for patients with Fabry disease. These data are supported by studies in adults in which ERT has limited ability to prevent end-stage organ damage when therapy is started in adulthood [54], and where ERT is most effective in patients who initiated treatment at a younger age with less renal involvement (10-year follow-up) [55].
Furthermore, whilst the clearance of glomerular endothelial cell inclusions and mesangial cell inclusions was observed in paediatric patients who underwent ERT irrespective of treatment regimen, the clearance of GL-3 from podocytes may be dose-dependent [47,50] as demonstrated by a significant correlation between cumulative agalsidase dose and GL-3 clearance in podocytes [47]. Patients receiving agalsidase 0.2 mg/kg EOW did not achieve GL-3 clearance from podocytes, whereas all patients receiving agalsidase beta 1.0 mg/kg EOW showed significant clearance [47]. Additionally, when agalsidase beta 1.0 mg/kg EOW was switched to agalsidase alfa 0.2 mg/kg EOW podocyte GL-3 scores increased [50]. Moreover, urine albumin levels seemed sensitive to ERT dose, with greater reductions observed with agalsidase 1.0 mg/kg EOW compared with agalsidase 0.2 mg/kg EOW [44,47]. The potential importance of ERT dose was further demonstrated by a reduction in GI outcomes and pain with agalsidase beta (1.0 mg/kg EOW), but increasing abdominal and neuropathic pain when the treatment was stopped, the dose of agalsidase beta was reduced, or when ERT was switched to agalsidase alfa (0.2 mg/kg EOW) [50].
Limitations of this systematic literature review are described elsewhere [16]. The analysis of published literature on ERT in paediatric patients with Fabry disease must take into account the heterogeneity of the patient population and the limited clinical evidence available for various outcomes. Larger studies, with age-matched healthy or untreated children would be needed to demonstrate any effects of ERT on delay to Fabry disease progression. Furthermore, it may be interesting to investigate the usefulness of urinary GL-3 and lyso-GL-3 as a noninvasive, easily measurable biomarker for monitoring paediatric patients with Fabry disease [35,[56][57][58].
In conclusion, ERT has a beneficial clinical impact in paediatric patients with Fabry disease, as it can relieve neuropathic pain, ameliorate gastrointestinal symptoms and improve quality of life. Moreover, ERT has been shown to normalize plasma GL-3 levels and to clear GL-3 inclusions from renal cells, with a dose-dependent mechanism in podocytes. The evidence of a clinical and of a metabolic response following treatment in young patients suffering from Fabry disease indicate that ERT may be important to prevent disease progression and end-stage organ damage.