Genotype‐phenotype relationships in mucopolysaccharidosis type I (MPS I): Insights from the International MPS I Registry

Abstract Mucopolysaccharidosis type I (MPS I) is a rare autosomal recessive disorder resulting from pathogenic variants in the α‐L‐iduronidase (IDUA) gene. Clinical phenotypes range from severe (Hurler syndrome) to attenuated (Hurler‐Scheie and Scheie syndromes) and vary in age of onset, severity, and rate of progression. Defining the phenotype at diagnosis is essential for disease management. To date, no systematic analysis of genotype‐phenotype correlation in large MPS I cohorts have been performed. Understanding genotype‐phenotype is critical now that newborn screening for MPS I is being implemented. Data from 538 patients from the MPS I Registry (380 severe, 158 attenuated) who had 2 IDUA alleles identified were examined. In the 1076 alleles identified, 148 pathogenic variants were reported; of those, 75 were unique. Of the 538 genotypes, 147 (27%) were unique; 40% of patients with attenuated and 22% of patients with severe MPS I had unique genotypes. About 67.6% of severe patients had genotypes where both variants identified are predicted to severely disrupt protein/gene function and 96.1% of attenuated patients had at least one missense or intronic variant. This dataset illustrates a close genotype/phenotype correlation in MPS I but the presence of unique IDUA missense variants remains a challenge for disease prediction.


| INTRODUCTION
Mucopolysaccharidosis type I (MPS I) is a rare recessive lysosomal storage disorder caused by pathogenic α-L-iduronidase (IDUA) gene variants which lead to deficiency of the lysosomal enzyme IDUA, EC 3.2.1.76. 1 Historically, MPS I patients have been classified into three phenotypic disease categories; Hurler, Hurler-Scheie and Scheie syndrome. 2 This nosology was based on age of symptom onset and presence of progressive intellectual disability; with Hurler syndrome representing the most severely affected individuals, characterized by onset of symptoms in early infancy with evidence of early progressive intellectual decline and when untreated, death within the first decade.
Hurler-Scheie and Scheie syndromes represent later onset of disease symptoms, more slowly progressive disease with sparing of intelligence. Although Hurler syndrome represents a distinctive phenotype, Hurler-Scheie and Scheie syndromes encompass a broad phenotypic spectrum with less precise delineation. From the practical perspective of disease management, a binary classification of patients into the categories of severe disease (Hurler syndrome) or attenuated disease (Hurler-Scheie and Scheie syndromes) is most useful. This classification facilitates management decisions based on the currently approved treatments for MPS I: hematopoietic stem cell transplantation (HSCT) and enzyme replacement therapy (ERT) with laronidase. 3 ERT has been shown to alter somatic disease symptoms and progression but as current recombinant enzyme does not cross the bloodbrain barrier this treatment approach does not address the progressive central nervous system (CNS) disease characteristic of Hurler patients. 4 HSCT particularly when initiated early, has been shown to preserve cognition and lead to improved developmental outcomes and is thus considered the standard of care for patients with severe MPS I. 5,6 The recent initiation of newborn screening for MPS I as well as other programs to identify individuals with MPS I at an age when CNS and somatic involvement may be minimal, highlight the need for accurate delineation of patients so effective therapies can be initiated early. [7][8][9][10][11][12] Unfortunately, no currently available biochemical assessments allow for accurate classification of patients as either severe or attenuated. 13 Published diagnosis and management guidelines suggest a role for IDUA genotype in management decision making. 7

| Study population
Clinical and molecular genetic data of patients enrolled in the international MPS I Registry as of April 2017 were evaluated. Patients were included if they had (a) known diagnostic date and date of birth, (b) severity designation of Hurler, Hurler-Scheie or Scheie as reported by the individual centers, (c) standardized IDUA genotype data with at least 1 IDUA sequence variant. Neither the laboratory method used to determine the sequence variants nor whether parental testing confirmed the inheritance or phase of the variants is captured by the Registry. In addition, the Registry does not capture known IDUA polymorphisms. A total of 1007 participants were enrolled in the MPS I Registry as of April 2017 with IDUA variant data available for 556 (55%). Those with available genotype data did not vary substantially from those who did not (n = 451). Eighteen patients had only a single variant reported with 538 having two reported variants. For genotype/phenotype associations, patients designated as Hurler were considered "severe" and patients designated as Hurler-Scheie or Scheie were considered as "attenuated." The terms "unique variants" and "unique genotypes" are defined as occurring in only one patient within this registry dataset and do not imply that the variant has not been previously published.

| Patient demographics
Of the 556 eligible patients, 538 had two IDUA variants reported.
Demographic information listing the geographic region of the Registry sites and basic diagnostic information on these 538 patients are listed in Table 1. The majority of patients are from European or North American sites. Latin America is under represented in this dataset as participants from this region are currently undergoing updated consenting.
Among the eligible patients, 380 patients were classified as severe and 158 classified as attenuated. As expected, the mean age of diagnosis among severe patients (1.2 years) is at a younger age than attenuated (7.0 years).

| Statistical summary of variants and genotypes
A total of 1094 variants were reported among the 556 eligible patients, of these, 148 different IDUA variants were identified.
Seventy-four variants, representing 50% of all variants occurred only once. Eighteen patients (3%) had only one allele identified; 10 were classified as severe and eight classified as attenuated. Figure 1 and  (Table 3) with the remaining 47 patients (12.6%) having unique genotypes (listed in Table S1). Thirty-seven of the 84 unique genotypes identified in severe patients were compound heterozygous with two severe variants (as defined above) and were comprised of 34 separate severe variants (listed in Table S1). There were nine missense recurrent variants and one intronic recurrent variant that were exclusively seen in severe patients (Table 5A).

| Attenuated patients
There were 87 individual genotypes represented in the 158 patients with an attenuated phenotype and two reported alleles; 24 genotypes were recurrent in a total of 95 patients (60%) and 63 unique (40%) (

| A327P
The A327P variant was identified in a total of 30 patients (Table 6B).
Five patients were homozygous for this variant three classified as severe and two attenuated. All 20 of the patients where A327P was compound heterozygous with a severe allele were classified as severe.

| P496R
The P496R variant was noted in seven patients; six severe and one attenuated. The single attenuated patient was compound heterozygous with L18P and the severe patients were compound with Q70X

| Deletions within the signal peptide
Twenty-one patients had deletions within the signal peptide region as designated in Figure 3.

DATA ACCESSIBILITY
The data that support the findings of this study can be requested by