Consensus recommendations for diagnosis, management and treatment of Fabry disease in paediatric patients

Abstract Fabry disease (FD), a rare X‐linked disease, can be treated with bi‐monthly infusion of enzyme replacement therapy (ERT) to replace deficient α‐galactosidase A (AGAL‐A). ERT reduces symptoms, improves quality of life (QoL), and improves clinical signs and biochemical markers. ERT initiation in childhood could slow or stop progressive organ damage. Preventative treatment of FD from childhood is thought to avoid organ damage in later life, prompting a French expert working group to collaborate and produce recommendations for treating and monitoring children with FD. Organ involvement should be assessed by age 5 for asymptomatic boys (age 12‐15 for asymptomatic girls), and immediately for children diagnosed via symptoms. The renal, cardiac, nervous and gastrointestinal systems should be assessed, as well as bone, skin, eyes, hearing, and QoL. The plasma biomarker globotriaosylsphingosine is also useful. ERT should be considered for symptomatic boys and girls with neuropathic pain, pathological albuminuria (≥3 mg/mmol creatinine), severe GI involvement and abdominal pain or cardiac involvement. ERT should be considered for asymptomatic boys from the age of 7. Organ involvement should be treated as needed. Early diagnosis and management of FD represents a promising strategy to reduce organ damage, morbidity and premature mortality in adulthood.

FD was traditionally considered to be an adult disease, but it is now recognised that disease processes and symptoms start in infancy or early childhood. Early manifestations of classic FD in children include pain (dysesthesia), reduced or absent sweating (hypohidrosis or anhidrosis), corneal whorls (cornea verticillata), angiokeratoma and gastrointestinal (GI) discomfort. 4 Pain represents the most striking symptom in children with FD, but overall quality of life (QoL) is often considerably impacted and characterised by fatigue, anxiety, depression, and school absences. 5 Later, adults can experience progressive renal disease, and cardiac and central nervous system (CNS) complications which contribute to morbidity and early mortality. 6 FD can manifest with a range of severities and has a spectrum of phenotypes. In classic FD, the first symptoms, including chronic neuropathic pain and episodic severe pain crises, typically emerge during childhood. Symptoms such as fatigue, hypohidrosis, skin angiokeratomas, diarrhoea, abdominal pain, hypoacousia, tinnitus, and cornea verticillata are additional common early manifestations. Occult kidney injury may occur at a young age, including pathological albuminuria and glomerulosclerosis. Symptomatic organ complications typically emerge in young adult patients, including chronic kidney disease (CKD) progression to renal failure and left ventricular hypertrophy (LVH) associated with myocardial fibrosis and arrhythmias, 7 pulmonary involvement, 8 sudden deafness 9 transient ischemic attacks, strokes, and eventually premature death. Non-classic phenotypes of FD include later-onset forms with predominant cardiac involvement due to pathogenic GLA variants such as p.Phe113Leu or p.Asn215Ser, the most frequent ones in Caucasian subjects, or IVS4+919G>A, highly prevalent in Chinese-Taiwanese populations. Later-onset phenotypes are frequently under-recognised because they lack classic manifestations of FD, such as acroparesthesia, cornea verticillata or angiokeratoma. 6 Screening studies of high risk populations (eg, patients with left ventricular hypertrophy or on haemodialysis) have identified previously undiagnosed FD in adults aged from 30 years up. [10][11][12][13] Routine screening of these at-risk populations may identify new cases, allowing initiation of effective treatment.
The natural history and characteristics of FD in children has been clarified in recent years by studies of patient registries. Symptoms appear at a median age of 6 years in boys and 7-8 years in girls. 14,15 The age of symptom onset was higher in a survey of index cases with no known family history of FD (10.9 years in boys and 22.6 years in girls). 16 The non-specific nature of FD symptoms in children can significantly delay diagnosis in index patients. Indeed, the delay between symptom onset and diagnosis of FD has been reported as 13.7 years in males and 16.3 years in females. 16 The longer time to diagnosis in females arises in part from disease heterogeneity associated with an X-linked disease but also from the now-discredited assumption that females with GLA pathogenic variants were merely carriers of FD. 17 It is now known that females can occasionally have severe FD, similar to the classic FD phenotype, seen most commonly in hemizygous boys with dramatically decreased (<1%) or no AGAL-A activity. 18 In classic FD, the phenotype and natural disease course in females is mainly determined by the pattern of X-chromosome inactivation (XCI). 19 Severe classic phenotype of FD arises in females when XCI pattern is skewed towards the mutant GLA allele in a ratio of 80:20 or greater across tissues. Echevarria 20 Agalsidase beta (Fabrazyme, Sanofi-Genzyme) is approved for children and adolescents aged 8 years and older at a dose of 1.0 mg/kg. 21 To date, agalsidase beta is the only FDA-approved ERT for FD.
A recent Cochrane review of evidence from randomised controlled trials shows that, compared to placebo, ERT reduces plasma Gb 3 and improves pain-related QoL. 22 The Cochrane review was complemented by a pooled analysis of cohort studies, and showed that patients taking agalsidase beta had a significantly lower incidence of cardiovascular, renal and cerebrovascular events than patients not taking ERT. 23 Various open-label studies have showed the efficacy and safety of ERT in children: ERT treatment decreases Gb 3 accumulation in tissue, plasma and urine, [24][25][26] and improves pain 27 and GI symptoms, 24,28,29 as well as QoL, energy and activity levels. 26,29,30 A long-term study of ERT efficacy in adults reported a 10-year survival rate of 94% and most patients (81%) had no severe clinical events during this time. Younger patients (mean age at agalsidase beta onset: 25 years) demonstrated the most benefit from ERT, and renal decline was slower in those patients, suggesting that early initiation of ERT could slow or even prevent renal decline. 31 An additional drug, the pharmacological chaperone migalastat, was approved in the EU in 2016 and in the USA in 2018. Migalastat is a small-molecule inhibitor of AGAL-A which, when bound at subinhibitory concentrations, corrects the folding error of some missense GLA variants thereby facilitating normal trafficking of AGAL-A from the endoplasmic reticulum through the Golgi to the lysosome. 32 Once in the lysosome, the complex dissociates, leaving functional AGAL-A to degrade accumulated substrate. 33 In the phase III FACETS trial (NCT00925301), migalastat did not achieve its primary endpoint vs placebo. After 24 months of treatment, in a post hoc analysis of the modified intent to treat (ITT) population, migalastat was associated with stabilisation of kidney function and improvement of left ventricular mass index 34 in adult patients with FD and amenable GLA variants. 34 Migalastat, which indication is limited to adult Fabry patients with amenable GLA variants, is not currently approved in children under the age of 16 in the absence of paediatric data.

| Existing guidelines
Although several guidelines and consensus recommendations exist for FD [35][36][37][38] there has been a shift in mindset over the past decade. This was prompted by better understanding that early ERT initiation can prevent organ damage in later life, improving both morbidity and mortality. The philosophy has changed from treatment to prevention, with the aim of maintaining organ function, optimising QoL and preserving life expectancy. There is consequently a need for updated recommendations for managing children with FD.
A French working group, comprising paediatricians and geneticists experienced in treating FD, convened three times between 2016 and 2017 to discuss the previous French guidelines 36 in light of developments in the field, and to discuss a management strategy for children with FD.
Here we present the proposals agreed by the group as a practical guide for diagnosis, treatment and monitoring of children with FD.

| DIAGNOSIS IN CHILDREN
FD is generally thought to be underdiagnosed. Recent screening studies report that prevalence of FD in males may be as high as 1:3100 39 and 1:8500. 40 These results are significantly higher than previous estimates of 1:117 000 41 and 1:476 000. 42 The estimate of 1:3100 38 included five newborns diagnosed with GLA variants which are not disease-causing (p.E66G, p.A143T-three cases, p.R118C). Exclusion of these patients reduces the incidence to an estimate of 1:5000 (DP Germain, unpublished data). Population-wide newborn screening is not currently implemented in any European country except Italy, while Taiwan and several states in the United States systematically screen newborn children for several lysosomal disorders including FD. 38,43 2.1 | Children from families with known Fabry disease Systematic screening of children with at least one first degree relative with FD is the simplest and fastest way to improve the rate of diagnosis. In France, FD in such children is usually detected by a geneticist or a paediatrician. In boys, the diagnosis is confirmed by nearly complete deficiency of AGAL-A activity in leukocytes, dried blood spot or plasma 36 along with confirmation of a pathogenic GLA variant. In girls, however, the diagnosis is more complex because XCI in heterozygotes can result in only partially deficient or even subnormal AGAL-A levels in the leukocytes and plasma. 19 Thus, the diagnosis of FD should always be confirmed by identification of a pathogenic GLA variant. 36,44 Of note, predictive genetic testing of children born to parents with FD is considered to be justified, according to the European Society of Human Genetics guidelines, because FD is now recognised to have onset in childhood and ERT could prevent damage in adulthood. 45 Prenatal diagnosis of FD is available in France for male, but generally not for female, foetuses. This discrepancy is for ethical reasons related to the heterogeneity of disease severity in females. Moreover, families with known FD can be offered pre-implantation genetic diagnosis (PGD) of embryos prior to implantation during assisted reproduction, but this remains rare.

| Diagnosis of index cases
Owing to the lack of specific symptoms, diagnosis of index cases of FD is usually delayed and rarely occurs during childhood. The main signs and symptoms of FD in childhood are pain, fatigue, GI problems, reduced or absent sweating, heat/cold and exercise intolerance and angiokeratoma. 14 Other symptoms include the presence of corneal whorls upon slit lamp examination, 46 hearing loss, 47 tinnitus, 48 pathological albuminuria, delayed growth and reduced QoL. The signs and symptoms that should alert paediatricians to the possibility of FD are presented in Table 1, along with their frequency and age of onset.
FD should also be considered (using cascade genetic testing) if any family members, especially males, died before age 50 due to stroke, cardiac arrhythmias, cardiac conductance problems or renal insufficiency. 36 A hallmark sign of classic FD in children is neuropathic pain in the hands and feet, most commonly in the palms, soles and fingertips. 49 Neuropathic pain has been reported by up to 72.3% of patients with FD 4,14,47 and is more frequently present in boys. The pain is most commonly described as "burning" followed by stabbing, tingling and shooting and triggered by physical exercise, heat and fever. 49 Patients can experience crises of excruciating pain that radiate proximally to the entire body and that can last up to several days, with no respite offered by analgesics. 49,50 The mean age of onset of pain has been reported as 10.1 years in boys 3 and 15 years in heterozygote girls 2 although one study reported median onset of neuropathic pain as 7 years in boys and 9 years in girls. 14 However, pain in children can also arise from diverse conditions such as growing pains. It can therefore be difficult for physicians to correctly attribute pain to a diagnosis of FD, especially in the absence of other symptoms or prior family history of FD that could alert physicians to the possibility of FD. Growing pains can be further differentiated from FD pain by assessing when and where the pain occurs: FD pain is most frequently experienced as burning, stabbing, tingling or shooting pain in the hands and feet and occurs during the day, particularly in late morning. In contrast, growing pains usually occur in the lower extremities during the late evening and night. 51 In addition to neuropathic pain, children with FD often experience GI symptoms such as abdominal pain and diarrhoea. 52 These symptoms are seen early in childhood with a median onset of 5 years in boys and 9.5 years in girls and constitute the initial symptoms of classic FD in up to 19% of patients. 13 Misdiagnosis of index cases is common with up to 25% of patients experiencing a previous misdiagnosis 16

| Announcing the diagnosis
For index cases, the diagnosis should be announced to the family according to Good Practice guidelines. 54 The disease is explained, and the family is supported for the next steps, including genetic counselling. 55 A detailed genealogy/family history is performed to screen/identify other cases in the same family. 55

| BASELINE CLINICAL WORK-UP AND SUBSEQUENT MONITORING
Following diagnosis of FD, a baseline assessment of organ involvement should be performed, followed by annual monitoring in boys and every 2-3 years in girls. A multi-disciplinary approach with specialists experienced in the care of FD patients is essential. The multidisciplinary team should include the following specialists: paediatrician, geneticist, pain specialist with intervention from other specialists as required. 36 Advice can be obtained from the national tertiary referral centre (www.centre-geneo.com).
For asymptomatic children from families with FD, we recommend that the baseline assessment of organ involvement occurs no later than 5 years of age for boys and 12-15 years of age for girls ( Figure 1).
Parents of asymptomatic FD children should also be educated to recognise the major signs and symptoms of FD in young children ( The procedures to be performed at baseline and subsequently at annual monitoring in boys or every 2 to 3 years in girls are presented in Table 2. In the case of later-onset phenotypes, the monitoring procedures should be focused on the affected organ (ie, heart) and its periodicity should be determined case by case. 56 In children diagnosed with pathogenic GLA variants responsible for later-onset phenotype, through either screening protocols or cascade genetic testing within families, particular attention should be given to appropriate genetic counselling and the delivery of detailed information on the natural course of this form. Because later-onset FD often become overt only in adulthood, patient monitoring should be adapted and unnecessary procedures and treatment avoided.

| Renal assessment
In general, children do not suffer from renal insufficiency until adulthood, when renal failure accounts for much of the morbidity and mortality associated with FD, particularly in males. 57 Accumulation of Gb 3 in kidney cells and podocytes foot process effacement can be observed in kidney biopsies from children with FD, even before pathological albuminuria and proteinuria manifest as early signs of renal involvement. 25,58,59 Performing routine renal biopsy in children remains an area of debate within the community of physicians involved in the care of patients with FD. 57 Renal biopsy has been proposed and shown to be safe by several authors 57-60 and should be considered in selected paediatric cases, particularly when the decision to start ERT is questioning or in children with significant proteinuria or decline in renal function; in whom renal biopsy is essential to rule out a second renal disease. 38 We recommend that albuminuria, proteinuria and GFR be assessed in all cases. In clinical practice, GFR can be estimated using the composite Schwarz formula/equation 61 rather than using the single classic markers of renal function. eGFR can be computed directly using height, serum creatinine, blood urea and serum cystatin C on https://www.kidney.org/professionals/KDOQI/gfr_calculatorPed. 62 The size of the kidneys should also be measured by ultrasound at the baseline assessment.

| Cardiac assessment
The major sign of cardiac involvement in FD is left ventricular hypertrophy (LVH), but conduction disturbance, such as short PR interval (due to accelerated conduction in the absence of accessory pathway), and rhythm disturbances, such us sinus bradycardia, have also been reported early in life. The benefit of ERT on conduction defects is currently unknown. 6

| Ophthalmological assessment
A complete ophthalmological assessment is recommended at the baseline assessment. The presence of cornea verticillata (vortex keratopathy or whorls) strongly supports the diagnosis of a classic Child from a family with Fabry Disease? Symptomatic? (see Table 1 Annual multi-organ monitoring (see Table 3) Symptomatic? (see Table 1

| Gastrointestinal assessment
At baseline and all monitoring visits, a GI assessment is recommended, along with establishment of a growth curve (using height and weight). The baseline audit of organ involvement should be performed following diagnosis for symptomatic children by age 5 in asymptomatic boys and by age 12-15 for asymptomatic girls.

| Metabolic assessment
Alpha-galactosidase A activity should be assayed at baseline. Measurement of urinary Gb 3 had being historically proposed 65

and had
shown an ERT dose-dependence response, 66 but its clinical relevance has not been elucidated. Consequently, plasma lysoGb 3 may prove a better biomarker of disease activity and its monitoring is currently preferred and recommended. 30,[67][68][69] . If ERT is initiated, antiagalsidase antibodies should be regularly monitored. 70

| Other assessments
QoL assessment using the paediatric Quality of Life Inventory (peds QL) 5  Moreover, careful consideration should be given to confirm or establish the pathogenicity of the identified GLA variant. 73 In particular, pLeu3Pro (DP Germain, unpublished data), p.Glu66Gln, p.

| Treating organ involvement
Pathological proteinuria can be treated with angiotensin enzyme converting (ACE) inhibitors or angiotensin receptor blocker (ARB). 35  is similarly divided for asymptomatic girls. In this paper, we recommend that ERT initiation should be considered for symptomatic boys and girls and that prophylactic ERT treatment be considered for asymptomatic boys since the age of 7 years old.
Estimates of the prevalence of FD have dramatically increased by around 10-fold, based on a recent Italian newborn screening study. 39 However, awareness of FD amongst medical professionals remains low. The diagnosis rate of FD must improve, given that specific therapies may protect major organs from progressive damage. There is an urgent need for more robust evidence for pain management strategies in children, and for long-term evidence of the protective effect of ERT when initiated in childhood. Solutions to alleviate the burden of biweekly infusions are also needed, and the role of anti-agalsidase antibodies needs to be further elucidated in various patients' population subsets.

| CONCLUSIONS
FD is a multisystemic disease that starts during childhood and worsens throughout adulthood. Boys with classic FD usually manifest with early onset symptoms impacting their quality of life during childhood. Girls usually have milder disease with later onset, but some girls may develop severe FD, similar to the phenotype observed in hemizygous boys. Early diagnosis and management of FD represent a promising strategy to reduce organ damage, morbidity and premature mortality in adulthood.
This paper provides expert consensus recommendations for the diagnosis, monitoring and treatment of children with FD ( Figure 1) and a basis for updating the guidelines to reflect recent developments in the field.

AUTHOR CONTRIBUTIONS
Each author participated in the three expert panel meetings, participated in establishing consensus recommendations, revised and approved the manuscript. The first author made substantial contribution to the development of this manuscript.