Natural History and Molecular Characteristics of Korean Patients with Mucopolysaccharidosis Type III

Background: Mucopolysaccharidosis type III (MPS III) is an autosomal recessive lysosomal storage disorder characterised by progressive neurocognitive deterioration. MPS III subtypes are clinically indistinguishable, with a wide range of symptoms and variable severity. The natural history of this disorder within an Asian population has not yet been extensively studied. This study investigated the natural history of Korean patients with MPS III. Methods: Thirty-four patients from 31 families diagnosed with MPS III from January 1997 to May 2020 in Samsung Medical Centre were enrolled. Clinical, molecular, and biochemical characteristics were retrospectively collected from the patients’ medical records and via interviews. Results: 18 patients had MPS IIIA, 14 had IIIB, and two had IIIC. Twenty (58.9%) patients were male. Mean age at symptom onset was 2.8 ± 0.8 years and at diagnosis was 6.3 ± 2.2 years. All patients with MPS IIIA and IIIB were classified into the rapidly progressing (RP) phenotype. The most common symptom at diagnosis was language retardation (88.2%), followed by motor retardation (76.5%), general retardation (64.7%), and hyperactivity (41.2%). Language retardation was more predominant in IIIA, and motor retardation was more predominant in IIIB. The mean age of the 13 deceased patients at the time of the study was 14.4 ± 4.1 years. The age at diagnosis and lag time were significantly older and longer in the non-survivor group compared with the survivor group (p = 0.029 and 0.045, respectively). Genetic analysis was performed in 24 patients with MPS III and identified seven novel variants and three hot spots. Conclusion: This study is the first to analyse the genetic and clinical characteristics of MPS III patients in Korea. Better understanding of the natural history of MPS III might allow early diagnosis and timely management of the disease and evaluation of treatment outcomes in future clinical trials for MPS III.

Patients with MPS III develop normally during the first year of life. Neurodegenerative symptoms occur due to accumulation of HS within the central nervous system (CNS), resulting in cognitive decline, motor dysfunction, and behavioural abnormalities [11]. Somatic effects such as facial coarseness, hearing loss, hepatosplenomegaly, and dysostosis multiplex are less pronounced than in other forms of MPS [12,13]. MPS III subtypes are clinically indistinguishable, with a wide range of symptoms and variable severity. MPS III can be further divided into rapidly progressing (RP) and slowly progressing (SP) phenotypes.
There is currently no effective treatment for MPS III, and only conservative management is possible. Disease-specific treatments for MPS IIIA and IIIB are being studied, including intrathecally delivered enzyme replacement therapy (ERT), substrate reduction therapy (SRT), hematopoietic stem cell transplantation, and gene therapy [7,[14][15][16]. A few studies on the natural history of MPS III have been reported in Caucasian populations [8,12,17]. However, the natural history of this disorder within an Asian population has not yet been extensively studied. In this study, we investigate the natural history of MPS III in Korean patients and the clinical differences between IIIA and IIIB of the RP phenotype and between surviving and deceased groups.

Materials and Methods
Thirty-four patients from 31 families with MPS III diagnosed from January 1997 to May 2020 in Samsung Medical Centre were enrolled in this study. All patients were diagnosed with MPS IIIA, IIIB, or IIIC by enzymatic activities in leukocytes and/or fibroblasts. In most cases, identification of pathogenic mutations was performed. Clinical, molecular, and biochemical characteristics were retrospectively collected from patients' medical records and via interviews. Multiple answers were permitted during the interview. One investigator conducted the database analysis and the face-to-face or telephone interviews. Demographic information including gender, age at diagnosis, age at the first seizure, age at death, cause of death, molecular and laboratory results, ambulatory status at the time of the latest medical records, physical examinations, and surgical interventions were collected. Any available results of the following investigations were also collected: brain magnetic resonance imaging (MRI), echocardiography, abdominal ultrasonography, and other investigations relevant to the disease. The Four-Point Scoring System (FPSS) [18] was administered as an MPS III-specific parent-completed disability questionnaire that rates motor, expressive language, and cognitive function on a 3-point scale (normal = 3, beginning regression = 2, severe regression = 1, and lost skill = 0). The interviews and scoring assessments were all conducted by the same interviewer. The phenotype of MPS IIIA and IIIB is determined by age at diagnosis before (RP) or after (SP) age 6 years, which is a reliable indicator of disease severity [1,8] and/or known genotype severity. The RP phenotype with MPS IIIC includes loss of speech and motor ability, generally occurring within 1-5 years of each other [19]; the SP phenotype progresses more slowly, as one could anticipate by its designation. Inability to associate three words at 15 months was noted as language retardation according to the Centers for Disease Control and Prevention (CDC) language milestones [20]. Motor retardation was defined as inability to walk alone at age 15.3 months or older (95th percentile) according to gross motor milestones of the World Health Organization (WHO) [21].

Statistical Analysis
To compare continuous variables, a parametric independent two-sample t-test was conducted. The Mann-Whitney U-test was used for nonparametric variables based on the normality assumptions from the Shapiro-Wilk test. The chi-square test or Fisher's exact test was used to compare categorical variables. The Kaplan-Meier survival curve was used for the survival rate of patients with MPS III and was compared between groups with the use of the log-rank test. Analysis was performed using R Statistical Package (version 3.6.2; Institute for Statistics and Mathematics, Vienna, Austria). A p-value < 0.05 was considered statistically significant.

Ethics Statement
The study was approved by the Institutional Review Board (IRB) of Samsung Medical Centre, Sungkyunkwan University School of Medicine, Seoul, Korea (IRB no. 2020-10-164-003). Informed consent was obtained from the patients and/or their parents.

Results
Data were collected from 34 Korean patients (58.9% males and 41.1% females) with MPS III from 31 families (Table 1). MPS IIIA, IIIB, and IIIC were present in 52.9% (n = 18), 41.2% (n = 14), and 5.9% (n = 2) of MPS patients, respectively. Enzyme activities in all patients were very low or undetectable, at less than 10% of the lower limit of the reference value. The mean age at symptom onset was 2.8 ± 0.8 years, and the mean age at diagnosis was 6.3 ± 2.2 years. The most common symptom at diagnosis was neurodevelopmental retardation. The most common initial symptom was language retardation (88.2%), followed by motor retardation (76.5%), both language and motor retardation (64.7%), and behavioural abnormalities (hyperactivity, aggression, and unawareness of dangerous situations) (41.2%) (Tables 1 and 2). In language milestones, only four children (11.8%) acquired the capacity to associate three words by age 15 months. The performance was not significantly different according to MPS III subtype, except patients with IIIB who started walking later than those with IIIA (p = 0.002). All patients with MPS IIIA or IIIB were diagnosed before the age of 6 years and/or associated with the severe genotype; therefore, all of them were classified into the RP phenotype. The mean diagnosis lag time (time interval between symptom onset and diagnosis) was 3.6 ± 2.5 years (Table 1).

Clinical Presentation
There were no differences in clinical characteristics between patients with MPS IIIA and IIIB except for age at first walking. The median age at first walking without assistance was 1.25 years (IQR, 1-1.37) in IIIA and 1.55 years (IQR, 1.42-1.6) in IIIB (Table 1). Language retardation in IIIA and motor retardation in IIIB were the most common initial symptoms. Patients with IIIB started walking later than patients with IIIA (p = 0.01) ( Table 2). In patients with MPS III, the most common accompanying symptoms after diagnosis were behavioural abnormalities and impairment of speech (Table 2). Frequent falls due to clumsy walking occurred in 91.2% of patients. The data included one case of a forehead laceration and two cases of brain haemorrhages due to falls. Hearing impairment was reported in 17 patients (50%), of whom seven used hearing aids. Thirty patients (88.2%) showed coarse facial features, hirsutism (85.3%), or Mongolian spots (64.7%). More than half (55.9 %) of the patients with MPS III were tested for heart disease. Mitral regurgitation, tricuspid regurgitation, and mitral valve thickening were detected in 78.9%, 84.2%, and 68.4% of respective patients, respectively, at the time of data collection. Among 21 patients who underwent abdominal ultrasonography, organomegaly was observed in 17 (80.9%) ( Table 1), and hernias were found in eight (23.5%). A box plot figure showing the clinical features of all 34 patients is presented as Supplementary Figure S1. Twenty-four patients (70.6%) underwent at least one surgical procedure. The most prevalent surgical interventions were gastrostomy (52%), tracheostomy (17%), herniorrhaphy (11%), tonsillectomy (11%), adenoidectomy (11%), and craniostomy (11%), followed by ear tube insertion (5%), scrotal hydrocelectomy (5%), and frenulotomy of tongue (5%) (Figure 1).   Neither sibling showed hyperactivity. The urinary GAG levels of the older and younger siblings were 502.5 and 984.4 mg GAG/g creatinine (reference range: <175 mg GAG/g creatinine), respectively, and elevated GAG was identified as heparan sulphate by thin-layer chromatography. Additionally, the HGSNAT activities of the siblings were 0.6 and 0.7 nmol/17 h/mg protein (reference range, 8.6-32 nmol/17 h/mg protein), respectively. The younger sibling was diagnosed with MPS IIIC at 2.3 years, while the older was diagnosed only at 5.1 years. The siblings showed the SP phenotype, with loss of speech and walking ability occurring after age 10 years.

Neurodegenerative Symptoms
All patients of MPS IIIA and IIIB were classified into the RP phenotype, showing rapid cognitive decline at a mean age of 4.4 ± 2.7 years. The clinical course of MPS IIIA and IIIB was tracked using the FPSS (Figure 2). The onset of speech regression (score of 2) was observed at a median age of 3.9 years (range, 2.1-5 years), whereas motor and cognitive functions began to regress at a median age of 5.4 years (range, 3-13 years) and 4.4 years (range, 1.5-11 years), respectively. Severe regression (score of 1) in speech, motor, and cognitive function was found at a median age of 7.3 years (range, 2-10 years), 10.4 years (range, 7-15 years), and 8.2 years (range, 5-13 years), respectively. Approximately 78.8 % of patients presented with epilepsy at a mean age of 10.0 ± 2.9 years. Convulsions were usually controlled with one or two antiepileptic drugs (AEDs). Age at loss of speech in MPS III subtypes is shown in Supplementary Figure S2A. Clumsiness in walking was In this study, two patients with MPS IIIC were siblings. The younger sibling [3] was allergic to milk formula and food, so she consumed breast milk exclusively for over one year and was diagnosed with vitamin D-deficient rickets. The patient showed mild coarse facial features, hepatomegaly, delayed motor development, and dysplastic vertebrae at 2 years. However, cognitive function and speech development were normal at that time. The older sibling showed delayed speech development and normal motor development. Neither sibling showed hyperactivity. The urinary GAG levels of the older and younger siblings were 502.5 and 984.4 mg GAG/g creatinine (reference range: <175 mg GAG/g creatinine), respectively, and elevated GAG was identified as heparan sulphate by thin-layer chromatography. Additionally, the HGSNAT activities of the siblings were 0.6 and 0.7 nmol/17 h/mg protein (reference range, 8.6-32 nmol/17 h/mg protein), respectively. The younger sibling was diagnosed with MPS IIIC at 2.3 years, while the older was diagnosed only at 5.1 years. The siblings showed the SP phenotype, with loss of speech and walking ability occurring after age 10 years.

Neurodegenerative Symptoms
All patients of MPS IIIA and IIIB were classified into the RP phenotype, showing rapid cognitive decline at a mean age of 4.4 ± 2.7 years. The clinical course of MPS IIIA and IIIB was tracked using the FPSS (Figure 2). The onset of speech regression (score of 2) was observed at a median age of 3.9 years (range, 2.1-5 years), whereas motor and cognitive functions began to regress at a median age of 5.4 years (range, 3-13 years) and 4.4 years (range, 1.5-11 years), respectively. Severe regression (score of 1) in speech, motor, and cognitive function was found at a median age of 7.  Figure S2B. Dysphagia began at a median age of 12 years (range, 5.4-16 years). At a mean age of 15.1 ± 1.9 years, the patients were in a fully bedridden state. Gastrostomy (18/34, 52.9%) was performed at a mean age of 17.2 ± 1.9 years, and tracheostomy was performed at a mean age of 17.8 ± 3.7 years. In siblings with MPS IIIC, the onset of speech regression (score of 2) was observed at a mean age of 10.5 ± 0.5 years. Motor functions and cognitive functions began to regress (score of 2) at the age of 11.5 ± 0.75 years and 11.0 ± 0.5 years, respectively.

Survival Analysis by Subtype
The mean follow-up period for MPS IIIA and IIIB was 12.6 ± 6.1 year. Of the 18 patients with MPS IIIA, six (33.3%) died; of the 14 patients with MPS IIIB, seven (50%) died. The mean age at death was 14.4 ± 4.1 years; these patients died of pneumonia (n = 8, 11.3-23.9 years), epilepsy (n = 1, 10.1 years), or of an unknown cause (n = 4, 11.4-21.4 years). When the survival rates between the MPS IIIA and IIIB groups were compared, there was no significant difference (p = 0.89) (Figure 3). Table 3 presents the comparison between the 13 patients who died and 21 survivors at the time of this study. Five patients (38.5%) who died were female, but the sex difference was not statistically significant. The age at symptom onset was similar between the groups, but the age at diagnosis was significantly older in the non-survival group (7.3 ± 1.8 years compared with 5.6 ± 2.3 years in the survivor group; p = 0.029). This indicates a significant increase in lag time to diagnosis in the non-survivor group (4.8 ± 3.0 versus 2.8 ± 1.8 years, p = 0.045). History of AED medication was significantly more frequent in the non-survivor group (100% versus 61.9%, p = 0.027).

Genetic Analysis
Genetic analysis was performed in 24 patients with MPS III. Most of the mutation types were missense mutations: 92.3% in SGSH mutations and 88.9% in NAGLU mutations ( Figure 4A). In addition, the NAGLU mutations were mostly located in exon 6 (83.3%) and the SGSH mutations were mostly located in exon 6 (50%) and exon 8 (50%) (Figure 4B

Discussion
This is the first study to describe the natural history and genetic features of MPS III in Korea. In this study, MPS IIIA was the most common type (52.9%), followed by type IIIB. IIIB is the most common in Taiwan (82.1%) [9] and in China (55.9%) [10], which is different from Korea. In the Asian population, there have been only two studies regarding the natural history of MPS III in China [10] and Taiwan [9]. These studies determined that speech delay and intellectual disability were the most prevalent clinical manifestations, followed by hyperactivity, similar to our cohort study. In the literature, no clinical differences in development were found between MPS IIIA and IIIB. In this study, we addition-

Discussion
This is the first study to describe the natural history and genetic features of MPS III in Korea. In this study, MPS IIIA was the most common type (52.9%), followed by type IIIB. IIIB is the most common in Taiwan (82.1%) [9] and in China (55.9%) [10], which is different from Korea. In the Asian population, there have been only two studies regarding the natural history of MPS III in China [10] and Taiwan [9]. These studies determined that speech delay and intellectual disability were the most prevalent clinical manifestations, followed by hyperactivity, similar to our cohort study. In the literature, no clinical differences in development were found between MPS IIIA and IIIB. In this study, we additionally evaluated motor development using age at first walking without assistance and noted a difference between the MPS IIIA and IIIB groups.
The first stage of MPS III, delay in speech and cognitive development, usually occurs between the ages of 1 and 3 years for RP phenotypes and around the age of 4 years for SP phenotypes [17]. In RP phenotypes, the first stage is immediately followed by a period of progressive neurocognitive decline, emergence of behavioural difficulties, and sleep disturbances. Patients with MPS IIIA and IIIB usually die within the first 2 decades of life. MPS IIIC has an onset in early childhood, with death generally occurring before the third decade [19]. Of the four subtypes, the RP phenotype is most frequently reported in MPS IIIA [13]. MPS IIIB is quite similar to IIIA, but more attenuated patients with this phenotype have been reported [22]. In our cohort, all patients presented with the RP phenotype, and there were no significant differences in survival rates between patients with IIIA and those with IIIB ( Figure 3). However, language retardation was more predominant in IIIA, and motor retardation was more predominant in IIIB (Table 2). There was a significant difference in motor retardation (p = 0.01) ( Table 2). MPS IIIA patients manifested epilepsy before age 10, whereas this occurred at a later age for patients with MPS IIIB. Epilepsy was usually well controlled with one or two AEDs in most patients. The younger sibling with MPS IIIC was diagnosed at the age of 2.3 years, which was much earlier than most diagnoses, even considering the wide range in the average age at diagnosis with IIIC (4.5 to 12 years) [8,[23][24][25] and the known mean age at symptom and sign onset (3.5 years) [6].
A total of 13 patients died during the study period, and there was a diagnostic delay of 7.3 ± 1.8 years in the mean age at diagnosis. The mean age at death for our patients with IIIA and IIIB was 14.4 ± 4.1 years, which is in line with previous studies of the natural history of MPS III [12,17,18,26,27] (Table 4). When the group of patients who had died at the time of the study was compared with the survivor group (Table 3), the mean age at symptom onset was similar between the two groups. Furthermore, the mean age at diagnosis for the survivor group was significantly younger (5.6 ± 2.3 years vs. 7.3 ± 1.8 years, p = 0.029), and the lag time to diagnosis in the survivor group was also significantly shorter than that of the non-survivor group (2.8 ± 1.8 years vs. 4.8 ± 3.0 years, p = 0.045). The delay in diagnosis of MPS III might be due to poor awareness caused by the rarity of the disease, the absence of minor somatic manifestations compared with the other types of MPS, and/or the main symptoms of neurologic regression (e.g., speech delay and behavioural problems) being similar to those of other neurologic disorders [28] such as idiopathic developmental delay, attention deficit hyperactivity disorder, and autism spectrum disorder [29][30][31]. In our cohort, 18 patients (100%) with IIIA and 11 patients (78.6%) with IIIB showed language retardation as an initial symptom. In addition, 11 (61.1%) patients with IIIA and 14 (100%) patients with IIIB showed motor retardation as an initial symptom. The median age at first walking was 1.25 years (IQR, 1-1.37) in IIIA and 1.55 years (IQR, 1.42-1.6) in IIIB (Table 1). Parents might tend to be more sensitive to language retardation than motor retardation, and a longer median diagnosis lag time was observed for MPS IIIB in our cohort (Table 1).
In this study, all patients with MPS IIIA and IIIB were classified into the RP type, and no mutations related to SP type were found. According to the Leiden open variation database of SGSH [32] and NAGLU [33], the total public variants reported for the SGSH gene are 217, and 112 were reported for NAGLU gene (last updated 17 September 2021), of which the majority are missense mutations. All novel variants in this study were predicted to be pathogenic by SIFT, Mutation Taster, and Polyphen-2 prediction software. The novel variants comprise 26.9% of the alleles for the SGSH gene and 16.7% of the alleles for the NAGLU gene (Table 5). Specific mutations in SGSH associated with the RP type are related to distinct geographical locations. p.Arg245His, p.Arg74Cys, 1091delCys, and p.Ser66Trp were the most frequent mutations in the Dutch (56.7%) [34], Polish (56%) [35], Spanish (45.5%) [36], and Italian (33%) [37] populations, respectively. In this study, p.Ser347Phe [38] and p.Asp235Asn [39,40] mutations were common (26.9% and 23.1% of alleles, respectively). The Asp235Asn mutation was reported to belong to the RP phenotype [40], whereas the Ser347Phe mutation has not been related to any phenotype [37,40]. p.Arg377Cys, p.Asn24Thrfs*240, p.Gln365Arg, p.Arg150Gln, and p.Asp235Asn [40] in SGSH ( Figure 5) belong to the RP type of MPS IIIA. p.Arg565Trp [46], p.Arg565Gln, and p.Arg565Pro have been reported as mutational hotspots in NAGLU. In this cohort, p.Arg565Pro accounts for 16.7% of all alleles detected, while p.Arg482Trp [43,47] accounts for 61.1% and is the hotspot [48]. p.Arg565Pro [43], p.Arg482Trp [42,49], p.Gln259*, and p.Arg203* in NAGLU were associated with the RP phenotype. Both p.Arg565Pro and p.Arg482Trp are located in exon 6, and all mutations found in exon 6 account for 83.3% of the total alleles. Therefore, exon 6 is assumed to be an important site in Korean patients with MPS IIIB. The IVS2+1G > A and p.Arg384* mutations in HGSNAT have been frequently reported globally [3,23,50,51].
Currently, several clinical trials are investigating different disease-modifying treatment options for MPS IIIA and IIIB. As a disease modifying treatment of MPS IIIA is imminent, early assessment of the phenotype (RP or SP), i.e., identification before progression of clinical signs and symptoms, is crucial both for the inclusion of patients in trials and for the assessment of treatment effects based on the expected natural history of the disease. De Ruijter et al. [52] showed that newborn screening for MPS IIIA was feasible by measuring HS concentrations or lysosomal protein concentrations in dried blood spots. However, the current lack of consensus regarding when to begin treatment is an ongoing problem, and anxiety is a common issue in parents due to positive screenings for pseudodeficiencies (false positive) or for variants of unknown significance (VUS).
This study has several limitations. First, this study was retrospective. Nevertheless, the amount of missing data was small, and the data were verified via telephone interviews and assessment of medical records. Second, the number of patients included in this study was small. However, this study was performed in a single centre, and our institution has managed most of the MPS population in Korea; therefore, our data reliably represent the general situation in Korea. All patients were managed by the same protocol, allowing consistent data analysis.

Conclusions
This is the first study that analysed the genetic and clinical characteristics of MPS III patients in Korea. MPS III should be considered when a patient shows neurodegenerative symptoms such as language retardation, motor retardation, or hyperactivity. MPS IIIA is the most predominant in Korea. There were no differences in clinical characteristics between patients with MPS IIIA and IIIB except for age at first walking. Early diagnosis is important to increase the survival rate for this population, and new therapeutic strategies will significantly change the course of the disease. Our findings can be used to develop quality of care strategies and provide guidance for clinical trial endpoint evaluations.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jpm12050665/s1, Figure S1: A box plot figure showing the clinical features of all 34 patients. Figure S2: (A-C) Age at speech loss, age at walking ability loss, and age at which gastrostomy was performed in MPS IIIA and IIIB. There were no significant differences between MPS IIIA and IIIB groups when using the log-rank test of the Kaplan-Meier survival curve.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: All data generated or analysed during this study are included in this published article.