Respiratory function during enzyme replacement therapy in late-onset Pompe disease: longitudinal course, prognostic factors, and the impact of time from diagnosis to treatment start

Abstract

Objective: To examine respiratory muscle function among late-onset Pompe disease (LOPD) patients in the Pompe Registry (NCT00231400/Sanofi Genzyme) during enzyme replacement therapy (ERT) with alglucosidase alfa by assessing the longitudinal course of forced vital capacity (FVC), prognostic factors for FVC, and impact of time from diagnosis to ERT initiation.

Methods: Longitudinal FVC data from LOPD (symptom onset > 12 months or ≤ 12 months without cardiomyopathy) patients were analyzed. Patients had to have baseline FVC (percent predicted upright) assessments at ERT start and ≥ 2 valid post-baseline assessments. Longitudinal analyses used linear mixed-regression models.

Results: Among 396 eligible patients, median baseline FVC was 66.9% (range 9.3-126.0). FVC remained stable during the 5-year follow-up (slope = - 0.17%, p = 0.21). Baseline FVC was lower among various subgroups, including patients who were male; older at ERT initiation; had a longer duration from symptom onset to ERT initiation; and had more advanced disease at baseline (based on respiratory support use, inability to ambulate, ambulation device use). Age at symptom onset was not associated with baseline degree of respiratory dysfunction. Differences between subgroups observed at baseline remained during follow-up. Shorter time from diagnosis to ERT initiation was associated with higher FVC after 5 years in all patients and the above subgroups using a cut-off of 1.7 years.

Conclusion: FVC stability over 5 years suggests that respiratory function is preserved during long-term ERT in real-world settings. Early initiation of alglucosidase alfa was associated with preservation of FVC in LOPD patients with better respiratory function at the time of treatment initiation.

Keywords: Alglucosidase alfa; Enzyme replacement therapy; Late-onset Pompe disease; Pompe disease; Registry; Respiratory function.

Fig. 1

FVC during ERT with alglucosidase alfa over time. a Percent predicted FVC in the upright position among all patients (N = 396). Individual lines represent each individual patient's FVC values plotted over time longitudinally. The thick dark line is the estimate of all patients from the mixed model adjusted for age at first ERT. The median follow-up time for all patients was 4.0 years (range 0.5–5.0; interquartile range 2.7–4.6). b Percent predicted FVC in the upright position by baseline FVC. FVCs for patients grouped by different baseline FVC classifications. Each line corresponds to patients classified into one of three groups by their baseline FVC %predicted as follows: ≤ 55%predicted; > 55 to < 80%predicted; or ≥ 80%predicted). Each line is the estimate from the mixed model adjusted for age at ERT initiation with their confidence intervals (shaded area around line). The group of patients with a baseline FVC ≥ 80 was the reference group. CI confidence interval, ERT enzyme replacement therapy, FVC percent (%) predicted forced vital capacity in the upright position, SD standard deviation

Fig. 2

Slopes of FVC for the Shorter-Time^a and Longer-Time^a groups among all patients and by subgroups. a Slopes of FVC for the Shorter-Time group. b Slopes of FVC for the Longer-Time group. ^aShorter-Time group includes patients with a time from diagnosis of Pompe disease to initiation of ERT with alglucosidase alfa ≤ the median (0–1.7 years); Longer-Time group includes patients with a time from Pompe diagnosis to initiation of ERT with alglucosidase alfa > the median (> 1.7 years). Estimates and p values were obtained from separate linear mixed effect regression models fitted with FVC as the outcome. Each model included years of follow-up, Shorter- vs. Longer-Time groups, and an interaction term between the Shorter- vs. Longer-Time groups and years of follow-up. Models are slightly different depending on the subgroups to avoid over-adjusting. Thus, further adjustments are described below. ^bThe mixed model was further adjusted for sex, baseline FVC, and age at ERT initiation. ^cThe mixed model was further adjusted for baseline FVC and age at ERT initiation. ^dThe mixed model was further adjusted for sex and age at ERT initiation. ERT enzyme replacement therapy; FVC percent (%) predicted forced vital capacity in the upright position, CI confidence interval; P is P value.

Fig. 3

FVC difference between Shorter-Time^a versus Longer-Time^a groups for all patients and by subgroup populations. ^aShorter-Time group includes patients with a time from diagnosis of Pompe disease to initiation of ERT with alglucosidase alfa ≤ the median (0–1.7 years); Longer-Time group includes patients with a time from diagnosis of Pompe disease to initiation of ERT with alglucosidase alfa > the median (> 1.7 years). Estimates and p values were obtained from separate linear mixed effect regression models fitted with FVC as the outcome. Each model included years of follow-up, Shorter- vs. Longer-Time groups, and an interaction term between the Shorter- vs. Longer-Time groups and years of follow-up. Models are slightly different depending on the subgroups to avoid over-adjusting. Thus, further adjustments are described below. The p value tested if the FVC difference between Shorter-Time vs. Longer-Time groups was statistically different. ^bThe mixed model was further adjusted for sex, baseline FVC, and age at ERT initiation. ^cThe mixed model was further adjusted for baseline FVC and age at ERT initiation. ^dThe mixed model was further adjusted for sex and age at ERT initiation. ERT enzyme replacement therapy, FVC percent (%) predicted forced vital capacity in the upright position, CI confidence interval, P is the P value