Society for the Study of Inborn Errors of Metabolism 2020 Virtual Symposium

A group presentation of the International Classification of Inherited Metabolic Disorders (ICIMD) was delivered by Dr Carlos Ferreira (National Institutes of Health, Bethesda, MD, USA), Professor Shamima Rahman (University College London, Great Ormond Street Institute of Child Health, London, UK) and Professor Johannes Zschocke (Innsbruck Medical University, Innsbruck, Austria). Professor Zschocke introduced this presentation by indicating that there have been several initiatives to classify inherited metabolic disorders; for example, the SSIEM Classification of Inborn Errors of Metabolism 2011 (updated in 2012); a proposed nosology of inborn errors of metabolism, which identified 1015 well-characterised disorders, published in 2019; and a proposal for a simplified classification of inborn errors of metabolism using a pathophysiological approach, also published in 2019, to act as a practical guide for clinicians.^1-3

Professor Zschocke next described the principle of the ICIMD. As stated on the website, the ICIMD is “a hierarchical, group-based collation of all currently known inherited disorders. It includes any primary genetic – mostly monogenic – condition in which alteration of a biochemical pathway is intrinsic to specific biochemical, clinical and/or pathophysiological features”.⁴ The purpose of the ICIMD is to “facilitate an improved understanding of the interconnections between conditions that share functional, clinical and diagnostic features”.⁴ In Professor Zschocke’s opinion, providing a unified general structure for classification of inherited metabolic disorders enables the ICIMD to be used in journals, seminars, teaching materials and textbooks, in addition to clinical decision-making, clinical studies and registries, and rare-disease initiatives.

Dr Ferreira next described the method for generating the ICIMD which first involved defining the term ‘inherited metabolic disorders’, and involved input from a large number of experts in the field.⁵ Professor Zschocke then highlighted the general structure of the ICIMID, which includes disorders of⁵:

• Intermediary metabolism (energy, nutrients)
• Lipid metabolism and transport
• Metabolism of heterocyclic compounds
• Complex molecule and organelle metabolism
• Cofactor and mineral metabolism
• Metabolic cell signalling.

Professor Zschocke next presented the interactive sunburst chart which is available at icimd.org. This chart allows visualisation of the different disorder groups and subgroups.^4,5 Clicking on each disorder group and subgroups opens the respective disorder tables with hyperlinks to other databases: disorder name (Orphanet database); gene name (HUGO Gene Nomenclature Committee database); OMIM number (Online Mendelian Inheritance in Man database); and IEMbase (Inborn Errors of Metabolism Knowledgebase).⁴ Dr Ferreira, Professor Rahman and Professor Zschocke then further described the disorder groups and subgroups as described by the ICIMD.

Dr Ferreira ended this presentation by outlining the future maintenance of the ICIMD. According to Dr Ferreira, the ICIMD will undergo continuous revision and update as suggested by the experts who participated in the consultation process, by international metabolic societies and other stakeholders, and by anyone within the metabolic community who is willing to make suggestions.

Industry-sponsored symposium by Amicus Therapeutics

Professor David G Warnock (University of Alabama at Birmingham Hospital, Birmingham, AL, USA) alongside a patient with Fabry disease described the diagnostic journey this patient had experienced due to a late diagnosis of Fabry disease. The purpose of this symposium was to help clinicians shorten the journey to a confirmed diagnosis of Fabry disease through education and increased awareness. The patient with Fabry disease described their long road to diagnosis, the symptoms they experienced, which started in childhood, their misdiagnoses and what led to their eventual diagnosis of Fabry disease. Additionally, as a result of their correct diagnosis, subsequent family members were also identified as having Fabry disease.

Throughout the symposium, Professor Warnock provided clinical information on Fabry disease as summarised below:

• Fabry disease is an X-linked lysosomal storage disorder characterised by variants in the GLA gene encoding the enzyme α-galactosidase A (α-Gal A).⁶
• Deficiency in the enzymatic activity of α-Gal A causes accumulation of globotriaosylceramide (Gb₃) and globotriaosylsphingosine (lyso-Gb₃) in various tissues and cells leading to tissue fibrosis and ultimately organ failure.^7-9
• On average, patients with Fabry disease may visit 10 different specialists before a diagnosis is confirmed.¹⁰
• From the onset of symptoms, a diagnosis of Fabry disease may be delayed by approximately 15 years.¹¹

Professor Warnock concluded this symposium with his take-home messages:

• In his opinion, specialists are often too narrowly focused on their ‘favourite organ’ and fail to appreciate the signs of a patient with multi-organ involvement due to a single disease process.
• Following a diagnosis of Fabry disease, a detailed baseline evaluation of organ damage should be performed.
• A definitive diagnosis of Fabry disease requires genetic testing.
• Family screening for Fabry disease is important.

Industry-sponsored symposium by Takeda

What can GLA variants tell us and what can’t they?
Dr Gheona Altarescu (Medical Genetics Institute, Shaare Zedeck Medical Center, Jerusalem, Israel) began her presentation by stating that the human genome contains millions of genetic variants that make each person unique. She explained that the term ‘variant’ can be used to describe an alteration which may be benign, pathogenic or of unknown significance; the term ‘mutation’ is also increasingly being replaced by ‘variant’.¹² Variants can be further classified as pathogenic, likely pathogenic, uncertain significance, likely benign or benign.¹³ Dr Altarescu also explained the term ‘copy number variant’, which “refers to the genetic trait involving the number of copies of a particular gene present in the genome of an individual. Copy number variants account for a significant proportion of the genetic variation between individuals”.¹⁴ Copy number variants have been detected in genes associated with cardiomyopathy, including the GLA gene where variants are linked to Fabry disease.¹⁵ In Dr Altarescu’s opinion, it is recommended that copy number variants also be taken into consideration in patients with suspected Fabry disease. Dr Altarescu explained, in her own words, that variants of unknown significance are those which have not been shown in a laboratory or published in the literature. She indicated that multi-dimensional analyses are recommended to be performed to understand the clinical outcome of a variant of unknown significance; for example, using biological, clinical and epidemiological data, and family screening.¹⁶

Dr Altarescu noted that >1000 variants of the GLA gene have been linked to Fabry disease.^17,18 She then discussed the research of Germain and colleagues, published in 2020, which aimed to determine consensus phenotypic classification for previously unclassified GLA variants from the fabry-database.org.¹⁹ Dr Altarescu explained that the study used data from the Fabry Registry (sponsored by Sanofi Genzyme).¹⁹ Initiated in 2001, the Fabry Registry is an ongoing, international, multi-centre, observational programme for patients with Fabry disease.²⁰ The authors of the study hypothesised that the Fabry Registry includes several patients with the same genotype and therefore potentially offers a unique opportunity to assess patient phenotypes.¹⁹ Using the presence or absence of clinical characteristics associated with either classical or late-onset Fabry disease, Dr Altarescu explained that the authors developed a novel Fabry phenotype consensus classification system to facilitate the classification of GLA variants reported in the Fabry Registry.¹⁹

Dr Altarescu highlighted that 991 GLA variants were identified in this study; the most common variants were missense (61.0%) and frameshift (25.5%) variants. Missense variants are either pathogenic and associated with classical or late-onset Fabry disease, or are benign variants with no or unspecific Fabry disease-unrelated phenotypes. Frameshift variants are pathogenic variants associated with classical Fabry disease.¹⁹ Dr Altarescu next explained the five-stage iterative system which was used by Germain and colleagues in this study for GLA genotype‒phenotype classification¹⁹:

• Stage 1: preliminary phenotype classification of GLA variants which were unclassified on fabry-database.org and were reported in ≥5 male or female patients in the Fabry Registry.
• Stage 2: (re)classification of phenotypes, including new GLA variants.
• Stage 3: review of three GLA variants without Stage 2 consensus phenotype classification.
• Stage 4: final review of assigned phenotype classifications.
• Stage 5: Kaplan–Meier analysis of severe clinical event-free survival.

The main results of the study were then outlined by Dr Altarescu¹⁹:

• During Stage 1, 55 GLA variants were identified, of which 11 (20%) were defined as pathogenic. Of these, 10 out of 11 pathogenic variants were associated with classical Fabry disease and one with the late-onset phenotype.
• During Stage 2, 33 GLA variants were identified, 30 of which were considered pathogenic; 25 of these variants were attributed to classical Fabry disease and five to late-onset Fabry disease. Consensus was not achieved for three GLA variants during Stage 2: p.Asp322Glu, c.999+2T>C and p.Ala143Thr.
• During Stage 3, Germain and colleagues reclassified p.Asp322Glu to a pathogenic variant associated with late-onset Fabry disease, rather than the classical phenotype, and c.999+2T>C as a pathogenic variant linked to classical Fabry disease. Classification could not be established for p.Ala143Thr; however, the authors suggested that this variant is ‘likely benign’ but further research is required.
• Despite a final review of the assigned phenotype classifications during Stage 4, the GLA missense variant p.Ala143Thr remained a variant of uncertain significance.

Based on the new GLA classifications identified, 228 out of 1305 male patients (17.5%) included in the Fabry Registry with GLA variants previously unclassified on fabry-database.org could now be classified. The majority of patients were determined to have a pathogenic GLA variant associated with classical Fabry disease (n=171) and 57 patients had late-onset Fabry disease. Using Kaplan-Meier analysis in Stage 5, male patients with pathogenic GLA variants with a classical phenotype reported first severe events at a younger age than patients with late-onset Fabry disease. For females, 244 out of 1466 patients (16.6%) with previously unclassified GLA variants on fabry-database.org could now be classified. Most female patients were identified as having a pathogenic GLA variant associated with classical Fabry disease (n=210) rather than late-onset Fabry disease (n=34). During Stage 5, Kaplan-Meier analysis indicated that the reporting of first severe events in female patients with a classical or late-onset Fabry disease was less pronounced than male patients.¹⁹ In Dr Altarescu’s opinion, the strength of this study was the use of the large Fabry Registry database as it is difficult to study and understand the significance of GLA variants in patients from individual clinics.

Dr Altarescu then provided her recommendations to help understand the role of a variant of unknown significance in patients with Fabry disease. From her own experience, she recommended using markers that can help elucidate the genotype‒phenotype correlation in a patient; for example, biochemical markers (α-Gal A enzyme activity or levels of lyso-Gb₃), biopsies and radiological or ultrasonographic markers. She also noted that, in her experience, determining the level of X-chromosome inactivation in female patients with Fabry disease can be useful, in addition to family segregation of the GLA variant. Next, Dr Altarescu discussed X-chromosome inactivation in more detail. She explained that this phenomenon is the normal failure of expression of one of the two X chromosomes in females, whereby all but one of the genes are inactivated, apparently at random, and have no phenotypic expression.²¹ Dr Altarescu presented data from one study which showed that 16 out of 53 female patients with Fabry disease exhibited skewed X-chromosome inactivation of the GLA gene. X-chromosome inactivation was found to significantly impact the phenotype of Fabry disease in these patients. Compared with patients with wild-type alleles, patients with skewed X-chromosome inactivation favouring variants in the GLA gene had significantly lower α-Gal A enzyme activity (p<0.001), significantly more disease severity on the Mainz Severity Score Index (MSSI; p<0.001) and the DS3 (p<0.01). Patients with skewed X-chromosome inactivation favouring variants in the GLA gene were also more likely to exhibit deterioration of kidney function and progression of cardiomyopathy.²² In line with the authors, Dr Altarescu highlighted that these data suggested that X-chromosome inactivation has a significant impact on the phenotype and natural history of Fabry disease in females.²²

Dr Altarescu concluded her presentation with the following take-home messages, based on her clinical experience:

• The detection of a gene variant does not equal diagnosis of a disease; therefore, all variants should be carefully interpreted.
• Multi-dimensional analyses should be performed to determine the pathogenicity of gene variants, particularly variants of unknown significance.
• Understanding the pathogenicity of a variant may require use of several biomarkers and genetic markers.
• The same gene variant may not respond equally to different treatments; therefore, each new treatment should be assessed individually and follow-up of the treatment is necessary.

Working with phenotypic variability
Professor Derralyn Hughes (Royal Free London NHS Foundation Trust and University College London, London, UK) began her presentation by highlighting that wide phenotypic variation occurs in Fabry disease. From Professor Hughes’ clinical experience, phenotypic variability can also occur between family members with Fabry disease and even between families with the same GLA gene variant. In Professor Hughes’ opinion, phenotypic variability in Fabry disease is not simply just a result of reduced α-Gal A enzyme activity, but is also due to other genetic or environmental factors. She highlighted that female patients with Fabry disease may exhibit a different level of disease severity and phenotype compared with male patients. Professor Hughes indicated that, in her opinion, this suggests that Fabry disease displays a phenotypic continuum where some disease outcomes are predictable and others are more difficult to understand. Professor Hughes then presented data from a retrospective, multi-centre study of 596 patients with Fabry disease.²³ Professor Hughes highlighted that the findings from this study showed that age is an important predictor of disease progression in Fabry disease and also of heterogeneity. Although, Professor Hughes pointed out that this study also indicated that Fabry disease symptoms are heterogeneous, even within a given age range and between males and females.²³ In Professor Hughes’ opinion, although it is possible to predict that certain manifestations of Fabry disease will progress with age (e.g. deterioration of renal function or risk of left ventricular hypertrophy), it may still be difficult for clinicians to predict how Fabry disease will progress in individual patients. Due to the heterogeneity of the disease, Professor Hughes next indicated that this can impact event-free survival. For example, males with classical Fabry disease have a higher risk of developing any cardiac, cerebral or renal event compared with males with late-onset Fabry disease (hazard ratio [HR] 5.63, 95% confidence interval [CI] 3.17‒10.00; p<0.001), whereas this risk is much lower in patients with the late-onset phenotype in males versus females (HR 1.98, 95% CI 1.07‒3.69; p<0.05).²³ Professor Hughes questioned what is driving the heterogeneity in Fabry disease and how this can be managed clinically.

Professor Hughes next highlighted that different variants of the GLA gene in patients with Fabry disease lead to varying levels of α-Gal A enzyme activity, and subsequently different levels of substrate accumulation.²⁴ She presented a study which found that plasma and urinary levels of Gb₃ and lyso-Gb₃ and related analogues differed between patients with known pathogenic variants of the GLA gene and correlated with the disease phenotype.²⁴ In Professor Hughes’ opinion, these findings suggest that varying levels of lyso-Gb₃ and related analogues may underlie phenotypic heterogeneity or disease pathology in Fabry disease. She also noted that the likely natural history of Fabry disease and the likely disease progression could, in some patients, be predicted depending upon the presence of known variants of the GLA gene. For example, patients with late-onset Fabry disease with the GLA variant N215S (c.644A>G) typically develop disease manifestations which affect the heart and other organs, and develop symptoms at a later mean (standard deviation) age compared with patients with other GLA variants (males, 51.7 [14.5] vs 13.2 [12.6] years; females, 36.8 [21.6] vs 25.1 [16.4] years).²⁵ In Professor Hughes’ opinion, it is important that genotype‒phenotype correlations of Fabry disease are shared as this can be informative for clinicians who are presented with a patient who has a GLA variant that they have not seen previously.

Next, Professor Hughes discussed a study in which she and her colleagues developed an age- and gender-adjusted version of the MSSI which assesses the disease severity of Fabry disease.²⁶ The findings from this study indicated that disease severity was different between variants of the GLA gene. For instance, Professor Hughes noted that patients with the N215S (c.644A>G) variant had significantly lower severity scores compared with those with the R227X (c.679C>T) variant (p<0.001).²⁶ Professor Hughes reiterated that phenotypic heterogeneity can be observed between family members with Fabry disease, even between those with the same GLA gene variant. She presented data from a pedigree analysis of one family with Fabry disease which showed that levels of Gb₃, lyso-Gb₃ and disease severity differed between family members, and even between siblings who were close in age.²⁷

In Professor Hughes’ opinion, the age-dependent variation in Fabry disease is important for the late-onset phenotype. She noted that, in her experience, phenotypic variation in patients with late-onset Fabry disease is compressed into the later decades of life. Professor Hughes highlighted that phenotypic heterogeneity between patients with late-onset Fabry disease may subtly relate to the level of accumulated substrates. She indicated that, although the levels of lyso-Gb₃ are less in patients with late-onset Fabry disease compared with classical Fabry disease, the lifetime exposure to these substrates reportedly correlates with the level of left ventricular hypertrophy and renal dysfunction.^25,28 In Professor Hughes’ opinion, these findings validate the relevance of age and lyso-Gb₃ as pathological factors of phenotypic heterogeneity, even in late-onset Fabry disease. Taken together, Professor Hughes explained, in her own words, that in terms of phenotypic heterogeneity, different GLA variants can lead to varying levels of α-Gal A enzyme activity and subsequently different levels of Gb₃ and lyso-Gb₃. Professor Hughes went on to say that higher levels of these substrates may cause patients with Fabry disease to develop clinical manifestations earlier; however, in her opinion, this concept does not explain the heterogeneity of Fabry disease within families, which may instead link to other genetic and environmental factors. For example, X-chromosome inactivation in females impacts the natural history and phenotype of Fabry disease.²²

Next, Professor Hughes discussed differential response to treatment in male and female patients with Fabry disease and how this can be managed by clinicians. In one study, male patients were more likely than female patients to continue to exhibit residual organ involvement despite 4 years of Fabry disease-specific treatment, agalsidase alfa.²⁹ Professor Hughes explained that, in her opinion, these findings may relate to when treatment was initiated in males and females, particularly as female patients may be diagnosed with Fabry disease earlier via family screening once a male relative has been identified with the disease. In addition, Professor Hughes noted that the level of anti-drug antibodies in males with Fabry disease have been observed to be higher than females, which may be due to the lower residual enzyme activity of α-Gal A in male patients. For example, one study showed that serum-mediated inhibition to enzyme replacement therapy in treated male patients was significantly higher compared with treatment-naïve males with Fabry disease (p<0.001). Whereas, there was no significant difference between serum-mediated inhibition to enzyme replacement therapy in treated and treatment-naïve female patients with Fabry disease (p=0.1385). Moreover, levels of lyso-Gb₃ and disease severity were significantly associated with increasing inhibition to enzyme replacement therapy.³⁰ In Professor Hughes’ opinion, the development of anti-drug antibodies could affect how well male patients with Fabry disease respond to disease-specific therapies.

Professor Hughes next presented algorithms for the diagnosis and assessment of male and female patients with Fabry disease developed by Biegstraaten et al. (Orphanet J Rare Dis 2015) and Mehta et al. (QJM 2010).^31,32 In her own opinion, Professor Hughes indicated that it is important for a causal link between clinical manifestations and disease pathology to be drawn prior to a diagnosis of Fabry disease, and that clinicians perform a thorough differential diagnosis. She also noted that a multi-system and multi-professional approach to diagnosis and assessment should be undertaken in any patient suspected of Fabry disease. Using this comprehensive approach, in her own words, Professor Hughes indicated that this would enable clinicians to understand the level to which the GLA gene variant and the levels of Gb₃ and lyso-Gb₃ accumulation are causing the clinical manifestations in patients. In Professor Hughes’ opinion, it is important that patients with Fabry disease receive disease-specific therapy, if indicated, and supportive treatment which should be individualised to each patient. She highlighted that, in her experience, the heterogeneity of Fabry disease requires a heterogeneity of response to disease management by clinicians.

Based on her own clinician experience, Professor Hughes ended her presentation with the following conclusions:

• Diagnosis of Fabry disease is not solely based on genetic, biochemical or clinical findings, and instead requires an integrated approach.
• Clinicians should comprehensively investigate new GLA gene variants which have not been previously observed.
• It is important to account for heterogeneity in Fabry disease, both in patients with the same GLA gene variant and between family members.
• It is recommended that all patients with Fabry disease be individually assessed and a personal treatment plan be developed.