Introduction
Alpha-mannosidosis (AM) is a rare autosomal recessive lysosomal storage disorder with an estimated prevalence of one in 500,000–1,000,000 live births (Meikle et al
2004; Meikle et al
1999). Pathogenic sequence variants in the
MAN2B1 gene cause a reduction in the activity of lysosomal alpha-mannosidase, resulting in impaired glycoprotein degradation in the lysosomes, and ultimately, impaired cellular function and apoptosis (Borgwardt et al
2015; Thomas
2001). AM presents as a multi-systemic disease, characterised by immunodeficiency, hearing impairment, facial and skeletal abnormalities and mental retardation, among other manifestations (Malm and Nilssen,
2008).
Other than supportive care, the only treatment option currently available for AM is allogeneic haematopoietic stem cell transplantation (HSCT) from a human leukocyte antigens (HLA)-matched donor, which has a variable outcome and carries a serious morbidity and mortality risk (Mynarek et al
2012; Borgwardt et al
2014a).
Velmanase alfa is a recombinant human lysosomal alpha-mannosidase, developed as intravenous (IV) enzyme replacement therapy (ERT) for AM (Borgwardt et al
2013). In phase I/II trials, velmanase alfa was associated with a sustained decrease in serum oligosaccharides after 18 months of therapy (mean percentage change −89.9%,
P < 0.001) and achievement of an average improvement of 39 steps in the 3-minute stair climb test (3MSCT;
P = 0.004) (Borgwardt et al
2014b). Velmanase alfa treatment was subsequently evaluated in a phase III placebo-controlled randomised trial (NCT01681953; Borgwardt et al
2017 [submitted]). Here we present long-term outcomes in patients with AM treated with velmanase alfa.
Methods
Study design
This study is an integrated analysis of efficacy and safety outcomes in patients with AM who participated in velmanase alfa trials and received therapy for up to 4 years in follow-up clinical trial or compassionate use (CU) programme.
Analysis population and database generation
Individual patient data from phase I/II (Borgwardt et al
2013) and III trials and the subsequent rhLAMAN-07 (NCT01908712), rhLAMAN-09 (NCT01908725) and rhLAMAN-10 (NCT02478840) studies were integrated into a single database. rhLAMAN-07 and rhLAMAN-09 are ongoing clinical trials of once-weekly 1 mg/kg velmanase alfa treatment in patients from France, or from Poland and Norway, respectively, who previously participated in velmanase alfa trials. rhLAMAN-10 is a single-centre clinical trial of 1 mg/kg velmanase alfa in which patients who had previously participated in velmanase alfa clinical trials and subsequently enrolled in the international CU programme were invited to undergo a comprehensive evaluation visit (last observation; LO). Inclusion/exclusion criteria are provided in the
Supplementary methods.
Patients had a confirmed diagnosis of AM, as defined by alpha-mannosidase activity <10% of normal activity, who had participated in the phase I/II and III trials, and were currently receiving weekly IV infusions of velmanase alfa according to their respective follow-up studies or CU programmes.
Procedures and treatment
All patients underwent clinical, functional and laboratory assessments at baseline and at pre-specified time points according to the protocol of their parental trial. Patients enrolled in rhLAMAN-07, −09 and − 10 underwent a LO visit at the same central location. In the rhLAMAN-10 trial, patients were screened for eligibility on day 1 of the visit. Patients who provided informed consent underwent pre-infusion evaluations, and were given a single IV infusion of velmanase alfa 1 mg/kg on day 2.
Endpoints
Primary endpoints for this analysis were the change from baseline to LO in serum oligosaccharides, and change from baseline to LO in the 3MSCT. Serum oligosaccharides were measured by high pressure liquid chromatography (HPLC) with ultraviolet (UV) detection coupled with matrix-assisted laser desorption/ionisation-time of flight (MALDI-TOF) mass spectrometry. Functional capacity was further assessed using the 6-minute walk test (6MWT), forced vital capacity (FVC % predicted and l, measured by spirometry), and Bruininks-Oseretsky test of motor proficiency (BOT-2). Immunological status was assessed as serum immunoglobulin G (IgG) concentrations and presence of hypogammaglobulinaemia in patients enrolled in the phase III trial, and classified according to criteria reported in Suppl. Table
2. Level of disability during activities of daily living and health-related quality of life were assessed using the Childhood Health Assessment Questionnaire (CHAQ) and Euro QOL 5D 5 L (EQ5D5L; used in the phase III trial) surveys, respectively, and results will be presented separately. Treatment-emergent adverse events (TEAEs), adverse drug reactions (ADRs), infusion-related reactions (IRRs) and anti-drug antibodies (ADAs) were assessed throughout the trials.
Statistical analysis
Baseline data were derived from the original trial in which patients were enrolled. LO data were derived from the last non-missing values collected per the protocol of the original trial, or long-term follow-on trial. For patients randomised to placebo in the phase III trial, the baseline for all evaluations was the last non-missing value recorded immediately prior to initiation of active treatment after study completion. Integrated analysis data were evaluated in the overall population and in age subgroups: adults (aged ≥18 years) and paediatrics (aged <18 years). The absolute changes and percentage changes from baseline to each time point were analysed for all primary and secondary efficacy endpoints, using the paired t-test and presented with their P-value and 95% CI.
Discussion
In this integrated analysis of the long-term efficacy of velmanase alfa treatment in patients with AM, statistically significant improvements were observed in the co-primary endpoints: serum oligosaccharide levels and 3MSCT. Secondary endpoints evaluating endurance, pulmonary function and motor proficiency also showed improvements up to 48 months, which are particularly relevant in the context of a progressively worsening disease. The long-term safety and immunogenicity profile of velmanase alfa appears compatible with chronic administration of the drug.
This study is a prospective integrated data analysis of previous clinical trials with different designs (rhLAMAN-02, −03, −04, −07 and − 09 are open-label, single-arm; rhLAMAN-05 is randomised, double-blind, parallel-group). The integrated study design was developed to address the challenges of the rarity of AM and statistical analyses of small patient populations. The study protocol was written a priori and the statistical analysis plan designed before database lock. The reason for choosing this approach arises from the rarity of the condition and the unusual possibility to be able to collect treatment data for up to 4 years before marketing authorisation.
The presence of a control group limited to the 12-month phase III trial (Borgwardt et al
2018) is, at least partially, mitigated by the duration of the follow-up and the repeated assessments.
Intra and inter-rater administrative reliability was maximised by conducting all assessments at one site with standardised administrative guidelines, and the same personnel collected the data on subsequent visits. The results of this analysis clearly support the biochemical efficacy of velmanase alfa treatment in patients with AM; marked decreases were seen in serum oligosaccharide levels, and statistically significant increases in serum immunoglobulins were observed, with correction of hypogammaglobulinaemia in many patients. Since the accumulation of mannose-rich oligosaccharides is considered the causative mechanism of cellular dysfunction and hypogammaglobulinaemia in AM, alongside with oligosaccharide accumulation in lymphocytes, and is suspected to be the cause of the increase in rate and severity of infections in AM patients (Malm et al
2000), these changes are assumed to produce a therapeutic benefit (Malm et al
2014; Muenzer
2014). The decrease in the proportion of patients who had impaired or seriously impaired immunoglobulin levels supports the use of serum immunoglobulin as an additional biomarker of velmanase alfa activity. All treated patients benefited from an improvement of IgG in serum.
Performance in functional assessments can be influenced by developmental stage, understanding of instructions and willingness to cooperate, all of which can be problematic in paediatric and/or cognitively disabled patients. These challenges, combined with the wide age range of study patients (6–35 years), may partially account for inter- and intra-patient variability. Six patients presented with concomitant conditions, such as psychotic behaviour or knee pain, that have compromised their endurance tests. As a severity score is lacking in alpha-mannosidosis, patients’ disease burden was evaluated at baseline based on the CHAQ disability index (DI). The patient population included in the rhLAMAN10 study scored differently, ranging from severe to mild disability. A post-hoc analysis revealed how, mean changes from baseline to LO showed an improvement in all baseline CHAQ-DI score groups in serum oligosaccharides, 3MWT, 6MWT and percentage of predicted FVC. As of today based on the currently available data, no baseline characteristic can be considered a predictive factor for VA treatment outcome. A post-marketing registry study will help in broadening the understanding of the heterogeneity of the alpha-mannosidosis population and the response to treatment.
The 3MSCT was chosen as an advanced activities-of-daily-living measure as it causes greater stress to the musculoskeletal and cardiorespiratory system, and requires a greater range of motion and muscle strength, compared with level walking (Nightingale et al
2014). A clear improvement in 3MSCT was evident in the paediatric population, which is notable given the progressive physical deterioration typically experienced by patients with AM and provides evidence supporting the effect of velmanase alfa treatment. In some patients, improved 3MSCT was also associated with a decreased reliance on wheelchair use and other walking help or aids.
For many endpoints in this study, the observed improvements were most marked in the paediatric population. To determine whether the 3MSCT improvements in the paediatric population were driven by growth alone, exploratory analyses were conducted within the paediatric age groups of <12 years and 12–17 years (representative of age groups characterised in healthy children by a slower growth and performance development in childhood vs the pubertal growth typical of adolescence). Interestingly, mean improvements in 3MSCT were similar between the two age groups at LO: +10.6 steps/min (+28.5%) in patients <12 years and + 10.7 steps/min (+18.3%) in patients aged 12–17 years, and thus not proportional to the improvement rate expected by growth alone.
The positive results for the 6MWT are consistent with the results from the 3MSCT and also suggest a mobility benefit associated with velmanase alfa treatment. A greater increase in FVC (% predicted) observed in the paediatric is also notable given the poorer pulmonary function at baseline in paediatric patients (mean FVC [%] at baseline 79.6% vs 92.5% in paediatric and adult patients, respectively). The comparative improvements in motor and pulmonary function in those who start treatment as paediatric patients, compared with those who begin treatment as adults, suggest that patients may benefit more from treatment started early in the disease course. This observation is in line with previous studies of long-term outcomes of patients receiving IV ERT for lysosomal storage disorders, and supports that treatment of patients with such disorders is recommended and should start early, preferably pre-symptomatically, to obtain better long-term outcomes (Muenzer
2014; Gabrielli et al
2010; McGill et al
2010; Tylki-Szymanska et al
2012; Tajima et al
2013).
A reduction in serum oligosaccharides and an increase in IgG levels were also achieved in adults. Clinically, these data are highly relevant when considering the vulnerability of this patient population to infections that can cause significant morbidity. In addition, although the magnitude of the treatment effect in domains such as endurance was smaller in adults compared with paediatric patients, the observed level of improvement or stabilisation in adults across all domains is still clinically important because of the progressive nature of the underlying disease. Preservation of adequate pulmonary function and maintenance of endurance and mobility in adult subjects may lead to a better health-related quality of life than would be associated with the gradual deterioration typical of AM.
Velmanase alfa was generally well tolerated throughout the study. A conservative threshold of 1.4 U/ml (lower limit of detection of the assay) was set with regard to detecting ADAs. A limited number of patients developed ADAs, with no clear effect on the co-primary efficacy outcomes (see also Borgwardt et al
2017), despite for one of the two patients with ADA > 80 U/ml who presented oligosaccharides above baseline levels at LO. Three patients experienced IRRs, all of which were mild or moderate in intensity and which resolved either spontaneously or with medical management. Of note, two of these patients had high levels of ADAs (>80 U/ml). A clear correlation between ADA and IRR occurrence has never been established in other therapies. Interestingly, the two patients with highest titres of ADA in VA clinical development programme developed IRR, suggesting the importance of keeping ADA monitoring in the future. The limited number of patients with high titre of ADA does not allow driving conclusions for today and the future registry study will help in providing new insights. The data were collected longitudinally, and form the largest clinical dataset evaluating ERT in AM to date. Our findings suggest that the significant improvements in biochemical and functional efficacy measures associated with velmanase alfa may persist for up to 4 years in paediatric patients, with adult patients experiencing significant improvements in serum oligosaccharide levels and stabilisation of functional performance up to 2 years after treatment initiation. The long-term safety outcomes suggest that there are no additional risks associated with extended treatment. The lack of effect on CSF biomarkers is an area of unmet need and a future research focus.
Compliance with ethical standards