Sleep disordered breathing in mucopolysaccharidosis I: a multivariate analysis of patient, therapeutic and metabolic correlators modifying long term clinical outcome

Abstract

Background: The lysosomal storage disorder, mucopolysaccharidosis I (MPS I), commonly manifests with upper airway obstruction and sleep disordered breathing (SDB). The success of current therapies, including haematopoietic stem cell transplantation (HSCT) and enzyme replacement therapy (ERT) may be influenced by a number of factors and monitored using biomarkers of metabolic correction. We describe the pattern of SDB seen in the largest MPS I cohort described to date and determine therapies and biomarkers influencing the severity of long-term airway disease.

Methods: Therapeutic, clinical and biomarker data, including longitudinal outcome parameters from 150 sleep oximetry studies were collected in 61 MPS I (44 Hurler, 17 attenuated) patients between 6 months pre to 16 years post-treatment (median follow-up 22 months). The presence and functional nature of an immune response to ERT was determined using ELISA and a cellular uptake inhibition assay. Multivariate analysis was performed to determine significant correlators of airway disease.

Results: The incidence of SDB in our cohort is 68%, while 16% require therapeutic intervention for airway obstruction. A greater rate of progression (73%) and requirement for intervention is seen amongst ERT patients in contrast to HSCT treated individuals (24%). Multivariate analysis identifies poorer metabolic clearance, as measured by a rise in the biomarker urinary dermatan sulphate: chondroitin sulphate (DS:CS) ratio, as a significant correlator of increased presence and severity of SDB in MPS I patients (p = 0.0017, 0.008). Amongst transplanted Hurler patients, delivered enzyme (leukocyte iduronidase) at one year is significantly raised in those without SDB (p = 0.004). Cellular uptake inhibitory antibodies in ERT treated patients correlate with reduced substrate clearance and occurrence of severe SDB (p = 0.001).

Conclusion: We have identified biochemical and therapeutic factors modifying airway disease across the phenotypic spectrum in MPS I. Interventions maximising substrate reduction correlate with improved long-term SDB, while inhibitory antibodies impact on biochemical and clinical outcomes. Monitoring and tolerisation strategies should be re-evaluated to improve detection and minimise the inhibitory antibody response to ERT in MPS I and other lysosomal storage diseases. Future studies should consider the use of sleep disordered breathing as an objective parameter of clinical and metabolic improvement.

Figure 1

Longitudinal trends in sleep disordered breathing (SDB) following treatment in MPS I patients treated with A. haematopoietic stem cell transplant (HSCT) or B. enzyme replacement therapy (ERT). Oxygen desaturation index 4% (ODI4%) data from sleep studies for each individual patient is plotted against duration since initiation of treatment, with solid lines connecting trends per patient. The shaded area represents an ODI4% of less than 10, representing the cut-off for severe SDB. Red legend identifies patients who require therapeutic intervention to the airway following the identified sleep study. (A). Data for 41 patients is presented. Hurler patients (circles) treated with HSCT show improvement in the severity of SDB compared to pre-treatment. Over the duration of follow up the majority of patients (76%) improve or remain stable. A reducing proportion of patients suffer severe SDB, with only one patient demonstrating severe SDB in the period after 3 years post HSCT. 3 patients require therapeutic intervention post-HSCT. (B). Data in 17 ERT treated attenuated patients (triangles) and 3 Hurler patients (circles) is presented. Amongst attenuated patients with longitudinal data, 73% are seen to progress, but rarely to severe SDB However, a cohort of attenuated patients developing inhibitory antibodies (open triangles) continue to suffer with severe SDB. 7 patients require therapeutic intervention (surgery or CPAP) while on ERT.

Figure 2

Metabolic biomarkers of clinical outcome in MPS I. Substrate reduction, measured by DS:CS ratio, and delivered enzyme activity following HSCT correlate significantly with measures of sleep disordered breathing (SDB) amongst MPS I patients. Bars represent means with p values presented from Mann–Whitney U (Multivariate p values are presented in Table 3). For correlation plots, linear regression lines of best fit were drawn and correlation coefficients were calculated with Pearson’s r, with p values representing significantly non-zero lines of best fit. Hurler patients identified by circle legend and attenuated by triangles. Each individual point represents one patient. (A). The ODI4% from the first post treatment sleep study of each patient correlates strongly to urinary DS: CS ratio performed at an identical time point (sleep study and urine sample collected within 4 weeks). Pearson Correlation r = 0.79 (r² = 0.68), p < 0.0001. (B). Mean urinary DS:CS ratio 1 year after treatment was significantly improved in individuals without SDB (mean 0.47, S.D 0.15), compared to those with SDB (mean 0.75, S.D 0.48, p = 0.04). Amongst ERT treated patients, patients with inhibitors drive worsening SDB. (C). Increasing iduronidase (IDUA) one year following transplant correlates significantly with improved DS:CS ratio at one year (r = −0.70, p < 0.001). (D). Iduronidase enzyme activity one year post transplant is significantly higher when SDB was absent (mean 44.69, S.D 20.8), compared to those where SDB was present (mean 29.33, S.D 12.9, p = 0.011). HSCT, Matched unrelated donor (MUD): closed circles. HSCT, Heterozygote donors: open circles. Attenuated patients on ERT without antibody response: closed triangles: Attenuated ERT patients with inhibitors: open triangles.

Figure 3

The role of inhibitory antibodies on metabolic biomarker and sleep disordered breathing (SDB) amongst ERT treated MPS I patients. Bars represent means with p values presented. For correlation plots, linear regression lines of best fit were drawn and correlation coefficients were calculated with Pearson’s r, with p values representing significantly non-zero lines of best fit. HSCT treated Hurler patients: closed circles. Hurler patients treated with ERT: Half shaded triangles. ERT patients with clinically significant inhibitory antibodies were defined as a cellular uptake inhibition greater than 30% & titres greater than 4000 (Open triangles). ERT treated patients without such an antibody response represented by closed triangles. (A). Correlation plot with linear regression lines of best fit for cellular uptake inhibition and DS:CS ratio. Increasing activity of inhibitory antibodies correlates strongly with poorer substrate clearance (urinary DS:CS ratio performed within 6 months of antibody status) (r² = 0.74, p = 0.002). 30% uptake inhibition is a suggested percentage with a measurable biochemical effect based on intersection of upper confidence intervals at baseline and lower CI. (B). Urinary DS:CS ratio within 6 months of antibody status was significantly improved in individuals without inhibitory antibodies (mean 0.61, S.D 0.14) compared to those with inhibitors (mean 1.8, S.D 0.76). As a reference, ERT attenuated patients without inhibitors have as successful metabolic outcomes as HSCT treated Hurler patients (mean 0.47). (C). SDB as measured by ODI4% within 6 months of antibody status was significantly lower in individuals without inhibitory antibodies (mean 2.03, S.D 1.68) compared to those with inhibitors (mean 16.85, S.D 7.23).