Eliglustat compared with imiglucerase in patients with Gaucher's disease type 1 stabilised on enzyme replacement therapy: a phase 3, randomised, open-label, non-inferiority trial

Articles

Eliglustat compared with imiglucerase in patients with Gaucher’s disease type 1 stabilised on enzyme replacement therapy: a phase 3, randomised, open-label, non-inferiority trial Timothy M Cox, Guillermo Drelichman, Renata Cravo, Manisha Balwani, Thomas Andrew Burrow, Ana Maria Martins, Elena Lukina, Barry Rosenbloom, Leorah Ross, Jennifer Angell, Ana Cristina Puga

Summary Background The mainstay of treatment for Gaucher’s disease type 1 is alternate-week infusion of enzyme replacement therapy (ERT). We investigated whether patients stable on such treatment would remain so after switching to oral eliglustat, a selective inhibitor of glucosylceramide synthase. Methods In this phase 3, randomised, multinational, open-label, non-inferiority trial, we enrolled adults (aged ≥18 years) who had received ERT for 3 years or more for Gaucher’s disease. Patients were randomly allocated 2:1 at 39 clinics (stratified by ERT dose; block sizes of four; computer-generated centrally) to receive either oral eliglustat or imiglucerase infusions for 12 months. Participants and investigators were aware of treatment assignment, but the central reader who assessed organ volumes was masked. The composite primary efficacy endpoint was percentage of patients whose haematological variables and organ volumes remained stable for 12 months (ie, haemoglobin decrease not more than 15 g/L, platelet count decrease not more than 25%, spleen volume increase not more than 25%, and liver volume increase not more than 20%, in multiples of normal from baseline). The non-inferiority margin was 25% for eliglustat relative to imiglucerase, assessed in all patients who completed 12 months of treatment. This trial is registered with ClinicalTrials.gov, number NCT00943111, and EudraCT, number 2008-005223-28. Findings Between Sept 15, 2009, and Nov 9, 2011, we randomly allocated 106 (66%) patients to eliglustat and 54 (34%) to imiglucerase. In the per-protocol population, 84 (85%) of 99 patients who completed eliglustat treatment and 44 (94%) of 47 patients who completed imiglucerase treatment met the composite primary endpoint (between-group difference –8·8%; 95% CI –17·6 to 4·2). The lower bound of the 95% CI of –17·6% was within the prespecified threshold for non-inferiority. Dropouts occurred due to palpitations (one patient on eliglustat), myocardial infarction (one patient on eliglustat), and psychotic disorder (one patient on imiglucerase). No deaths occurred. 97 (92%) of 106 patients in the eliglustat group had treatment-emergent adverse events, as did 42 (79%) of 53 in the imiglucerase group (mostly mild or moderate in severity). Interpretation Oral eliglustat maintained haematological and organ volume stability in adults with Gaucher’s disease type 1 already controlled by intravenous ERT and could be a useful therapeutic option. Funding Genzyme, a Sanofi company.

Introduction Gaucher’s disease is a rare metabolic disorder due to deficiency of acid β-glucosidase (β-glucocerebrosidase, GBA1; D-glucosyl-N-acylsphingosine glucohydrolase, EC 3.2.1.45). Glucosylceramide accumulation occurs in macrophages and patients present with hepatosplenomegaly, anaemia, thrombocytopenia, and bone disease; some patients have neurological manifestations.1 Gaucher’s disease type 1 is the non-neuronopathic form. Without treatment, Gaucher’s disease can shorten life and cause chronic disability.2 Alternate-week infusions of enzyme replacement therapy (ERT) with recombinant human acid β-glucosidase is the standard of care for Gaucher’s disease type 1; this treatment reverses the haematological and visceral complications of the disease, can prevent irreversible www.thelancet.com Vol 385 June 13, 2015

skeletal damage, and improves quality of life.2–6 ERT augments residual activity of acid β-glucosidase and promotes degradation of its endogenous substrates. Substrate reduction therapy (SRT) is an alternative that seeks to balance glucosylceramide production with its impaired rate of degradation by partly inhibiting glucosylceramide synthase, the rate-limiting step for glycosphingolipid biosynthesis.7 Miglustat is the only approved SRT for adults with Gaucher’s disease, but because of its less favourable efficacy, safety, and tolerability profile than ERT, has been licensed only for patients who are not suitable candidates for ERT.8–10 Eliglustat is a selective and potent inhibitor of glucosylceramide synthase, which has been approved by the US Food and Drug Administration as a first-line oral drug for treatment of adults with Gaucher’s disease type 1,7 and has also been

Lancet 2015; 385: 2355–62 Published Online March 26, 2015 http://dx.doi.org/10.1016/ S0140-6736(14)61841-9 See Comment page 2328 Department of Medicine, University of Cambridge, Addenbrooke’s Hospital, Cambridge, UK (Prof T M Cox FMedSci); Hospital de Niños ‘Dr. Ricardo Gutiérrez’, Buenos Aires, Argentina (G Drelichman MD); State Institute of Haematology ‘Arthur de Siqueira Cavalcanti’, Rio de Janeiro, Brazil (R Cravo MD); Mount Sinai Hospital, New York, NY, USA (M Balwani MD); Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA (T A Burrow MD); Universidade Federal de São Paulo, São Paulo, SP, Brazil (A M Martins MD); Hematology Research Center, Moscow, Russia (Prof E Lukina MD); Cedars-Sinai Oncology, and Tower Hematology Oncology, Beverly Hills, CA, USA (B Rosenbloom MD); and Genzyme, a Sanofi company, Cambridge, MA, USA (L Ross MD, J Angell MSc, A C Puga MD) Correspondence to: Prof Timothy M Cox, Department of Medicine, University of Cambridge, Addenbrooke’s Hospital, Cambridge CB2 0QQ, UK [email protected]

2355

Articles

granted marketing authorisation by the European Commission. Unlike miglustat, eliglustat resembles the ceramide rather than the glucose moiety of glucosylceramide and does not cross the blood–brain barrier.7,9,10 Because patients with Gaucher’s disease type 1 need life-long treatment, an effective, well-tolerated, oral treatment could provide a valuable first-line alternative to intravenous infusions every 2 weeks. In an open-label, phase 2 trial11–14 and a phase 3, randomised, double-blind, placebo-controlled trial in untreated adults,15 eliglustat reduced plasma glucosylceramide concentrations and improved visceral, haematological, and skeletal manifestations of Gaucher’s disease type 1. We describe the results of ENCORE, the largest prospective trial of any treatment in Gaucher’s disease type 1. Our aim was to establish whether patients who had achieved therapeutic goals while receiving alternate-week infusions of ERT would remain stable after switching to oral eliglustat.

Methods Study design and participants This randomised, multinational, open-label, noninferiority, active comparator, phase 3 trial compared eliglustat (previously Genz-112638; Genzyme Corporation, MA, USA) with imiglucerase (Genzyme, a Sanofi company). We enrolled patients at 39 clinics in Latin America, the USA, Canada, Australia, the Middle East, Europe, and Russia. Eligible patients were at least 18 years old, with non-neuronopathic Gaucher’s disease confirmed by enzymatic assay, and had received ERT for 3 years or longer, with a total monthly dose of 30–130 U/kg for at least 6 of the 9 months before randomisation. Patients needed to have achieved prespecified therapeutic goals for Gaucher’s disease: haemoglobin concentration of 110 g/L or more in women and 120 g/L or more in men; platelet count of 100 × 10⁹/L or more; spleen volume of less than 10 multiples of normal (MN; calculated by comparison with normal spleen volume, 0·2% bodyweight)16 or splenectomy 3 years or more before randomisation; liver volume of less than 1·5 MN (calculated by comparison with normal liver volume, 2·5% of bodyweight);16 and no bone crisis or symptomatic bone disease (bone pain attributable to osteonecrosis or pathological fractures within the last year). Exclusion criteria included use of miglustat within 6 months before randomisation; neurological complications, including peripheral neuropathy; pulmonary involvement; deficiency of iron, vitamin B12, or folate; clinically significant coronary artery disease, arrhythmias, or conduction defects; liver enzymes or total bilirubin more than twice the upper limit of normal (excluding Gilbert’s syndrome); pregnancy or lactation; and need for transfusion. A full list of inclusion and exclusion criteria are provided in the appendix. Because the cytochrome P450 (CYP) variants CYP2D6 and, to a lesser extent, CYP3A4 affect eliglustat breakdown, we genotyped patients for CYP2D6-predictive metaboliser 2356

status during screening. We monitored CYP2D6 inhibitors, CYP3A4 inhibitors, and CYP3A4 inducers, and permitted them for temporary or long-term use depending on CYP2D6 phenotype. We provided all patients with a list of prohibited drugs, including those known to prolong QTc interval. However, we made exceptions for temporary use of some QTc-prolonging drugs such as certain antibiotics after completion of the dose-adjustment period. We obtained written informed consent from each patient. The study was done in accordance with the Good Clinical Practice guideline defined by the International Conference on Harmonisation, principles defined in the Declaration of Helsinki and its amendments, and applicable national and international laws. The protocol was approved at each site by the institutional review board or independent ethics committee.

Randomisation and masking We stratified eligible patients according to ERT dose (<35 U/kg every other week or ≥35 U/kg every other week) and randomly allocated them 2:1 to receive eliglustat or imiglucerase for 12 months. The allocation sequence list, with blocks of four within each dose stratum, was computer-generated centrally by the funder before the start of the study. Patients were allocated their treatment group by the funder’s central clinical research pharmacy after receiving a request for randomisation from the study site physician. Participants and investigators were not masked to treatment assignment because enzyme therapy is an infusion and eliglustat is an oral capsule. However, the central readers from BioClinica, Lyon, France, who assessed organ volumes were masked to treatment allocation.

Procedures Patients randomly allocated eliglustat received 50 mg twice daily for 4 weeks. Thereafter, they received 50 mg or 100 mg twice daily until week 8, and 50 mg, 100 mg, or 150 mg twice daily, depending on the plasma concentration of eliglustat (≥5 ng/mL [≥12 nmol/L] or <5 ng/mL [<12 nmol/L]) at weeks 2 and 6. Thus, we individualised eliglustat dose on the basis of plasma concentrations. Patients randomly allocated imiglucerase received treatment every other week at their usual doses. We did the following assessments at baseline (two screening visits and day 1), 3 months, 6 months, 9 months, and 12 months: complete physical examination; vital signs; pregnancy test; urinalysis; routine biochemical tests, haemoglobin concentration and platelet count, biomarkers, and 12-lead electrocardiogram (ECG). We assessed spleen and liver volume with MRI and quality-of-life assessments at baseline, 6 months, and 12 months. We did Gaucher assessments (mobility, bone crises, and bone pain) and quality-of-life questionnaires at baseline, 6 months, and 12 months. We did bone analyses (femur and lumbar spine MRIs, dual x-ray absorptiometry, and spine x-rays), two-dimensional echocardiogram www.thelancet.com Vol 385 June 13, 2015

Articles

with Doppler, and complete neurological examinations, including nerve conduction studies and mini-mental state examinations, at baseline and 12 months. Additionally, patients allocated eliglustat had plasma drug monitoring and pregnancy testing at weeks 2, 4, 6, and 8, and months 3, 6, 9, and 12. Assessments were done by the investigators at their respective study sites. We continuously monitored adverse events and concomitant drug use.

Outcomes The composite primary efficacy endpoint was the percentage of patients whose haematological variables and organ volumes remained stable after 12 months according to the stability criteria established for these measures in patients with Gaucher’s disease type 1 receiving maintenance treatment with imiglucerase.17 These stability criteria were haemoglobin concentration that did not decrease by more than 15 g/L; platelet count that did not decrease by more than 25%; spleen volume (MN) that did not increase by more than 25%, and liver volume (MN) that did not increase by more than 20% from baseline. We also examined a binary (yes or no) composite endpoint on the basis of stable haematological and organ volumes around the reference range (appendix). Instances of failure to meet the primary endpoint were reviewed by an independent adjudication board masked to treatment. Likewise, imaging data were analysed by central readers masked to patient, treatment, and timepoint. Secondary efficacy endpoints included haemoglobin concentration, platelet count, spleen and liver volumes assessed by MRI, and dual-energy x-ray absorptiometry T scores and Z scores for femoral and lumbar spine bone mineral density. Additional endpoints included bone disease assessments (radiography, MRI, and bone marrow burden score);18,19 Gaucher’s disease assessments (mobility, bone crisis, and bone pain); quality-of-life assessments (brief pain inventory, fatigue severity scale, and 36-item short form health survey); Gaucher’s disease severity score; and preference for oral or intravenous treatment (questionnaire completed by all patients at screening and 12 months after treatment initiation for the eliglustat group). Biomarkers were analysed by central laboratories, and included plasma chitotriosidase (Labcorp, New Mexico, USA), macrophage inflammatory protein 1β (LabCorp Clinical Trials, California, USA), glucosylceramide, GM3, ceramide, and, sphingomyelin (Genzyme Clinical Specialty Laboratory, Massachusetts, USA). We assessed safety through monitoring of treatmentemergent adverse events, standard clinical variables, and recommended assessments for Gaucher’s disease, including nerve conduction studies. Adverse event severity, seriousness, and relatedness to treatment were assessed by the study site physician as per the clinical safety data management definitions and standards established by the International Conference on Harmonisation. The Data Safety Monitoring Committee, consisting of five members in the USA, Italy, and Germany, did periodic and ad-hoc www.thelancet.com Vol 385 June 13, 2015

reviews of safety data, including serious adverse events thought to be treatment related by the study site physician, and patient discontinuations or withdrawals. The Committee provided recommendations about clinical trial conduct and ethical guidance to ensure the safety and wellbeing of participating patients. Because data from preclinical studies (Bree M, Genzyme, personal communication) showed that eliglustat inhibits human potassium, sodium, and calcium cardiac ion channels in vitro, we also monitored patients electrocardiographically.

Statistical analysis This non-inferiority study was powered at 85% on the basis of a sample size of 105 or more evaluable patients at 12 months, a one-sided significance level of 0·025, a non-inferiority margin of 25% for eliglustat relative to imiglucerase, and a 20% non-evaluable or dropout rate. We based the non-inferiority margin on a 95% imiglucerase response rate and an 85% eliglustat response rate (established by phase 2 results).12 This margin is less than half the expected difference of 51% between imiglucerase treatment and discontinued treatment after 1 year with data from a matched population from the International Collaborative Gaucher Group Gaucher Registry (Angell J, unpublished). The margin also allows for a 10% difference between imiglucerase and eliglustat, and an additional 15% for inherent variability in estimation of the difference between these treatments.

See Online for appendix

209 assessed for eligibility

160 randomly allocated

106 allocated eliglustat

54 allocated imiglucerase

1 withdrew before treatment

106 received treatment

2 withdrew before 12 months 2 did not comply with ≥80% dose 2 allocated incorrect ERT dose strata 1 missing baseline or 12 month platelet count, or haemoglobin value

99 completed 12 months (per-protocol population)

53 received treatment

1 withdrew before 12 months 3 did not comply with ≥80% dose 2 allocated incorrect ERT dose strata

47 completed 12 months (per-protocol population)

Figure 1: Trial profile ERT=enzyme replacement therapy.

2357

Articles

Eliglustat (n=106)

Imiglucerase (n=53)*

Sex Male

47 (44%)

25 (47%)

Female

59 (56%)

28 (53%)

Race White

98 (92%)

48 (91%)

Black

6 (6%)

4 (8%)

Asian

1 (1%)

1 (2%)

White or American Indian

1 (1%)

0

Jewish descent

29 (27%)

14 (26%)

Age at start of study treatment (years)

37·6 (14·2)

37·5 (14·9)

Age at first Gaucher’s disease symptom (years)

12·7 (12·0)

15·7 (14·2)

Age at Gaucher’s disease diagnosis (years)

17·8 (13·6)

20·3 (14·3)

Number of years on imiglucerase (before randomisation)

9·8 (4·0; n=73)

10·0 (3·6; n=34)

ERT type at study entry Imiglucerase

84 (79%)

45 (85%)

Velaglucerase

22 (21%)

8 (15%)†

Baseline ERT dose <35 U/kg every 2 weeks

43 (41%)

22 (42%)

≥35 U/kg every 2 weeks

63 (59%)

31 (58%)

Gaucher genotype L444P/other

2 (2%)

N370S/L444P

38 (36%)

0 18 (34%)

N370S/N370S

23 (22%)

12 (23%)

N370S/other

34 (32%)

14 (26%)

Other/other

9 (8%)

9 (17%)

CYP2D6 metaboliser status Poor

4 (4%)

2 (4%)

Intermediate

12 (11%)

9 (17%) 39 (74%)

Extensive

84 (79%)

Ultrarapid

4 (4%)

1 (2%)

Indeterminate

2 (2%)

2 (4%)

1·15 (1·31)

1·12 (0·95)

Acid β-glucosidase activity (nmol/h per mg leukocyte protein)

Role of the funding source

Splenectomy No

76 (72%)

44 (83%)

Partial

1 (1%)

1 (2%)

Total

29 (27%)

8 (15%)

Spleen volume (MN)

3·17 (1·35)

2·74 (1·15)

Liver volume (MN)

0·94 (0·19)

0·92 (0·16)

Haemoglobin concentration (g/L)

136 (13)

139 (13)

Platelet count (×10⁹/L)

203 (79)

188 (57)

Chitotriosidase activity (nmol/h per mL) Total BMB score

1159 (1465)

1105 (1059)

8·22 (2·66)

8·12 (2·63)

Lumbar spine BMD T score

–0·54 (1·38)

–0·34 (1·15)

Femur BMD T score

–0·15 (1·09)

–0·41 (1·28)

Data are n (%) or mean (SD). ERT=enzyme replacement therapy. CYP=cytochrome P450. MN=multiple of normal. BMB=bone marrow burden. BMD=bone mineral density. *One patient who withdrew before treatment is not included. †We switched patients taking velaglucerase to imiglucerase at trial onset.

Table 1: Baseline characteristics

As is customary for non-inferiority studies, we did the primary efficacy analysis on the per-protocol population.20,21 This analysis included patients allocated treatment who had 80% or higher treatment compliance, 2358

no major protocol deviations that might interfere with efficacy, and no haematological decline from disorders other than Gaucher’s disease type 1. Patients receiving eliglustat who returned to imiglucerase were included in the per-protocol analysis and were deemed treatment failures. We also did all efficacy and safety analyses in all randomly allocated patients who received at least one dose of eliglustat or imiglucerase. We calculated the percentage of patients remaining stable and the exact 95% CI 12 months after treatment was started for both groups. We calculated the exact 95% CIs for each of the two randomisation dose stratification groups. We calculated the difference in the percentage that remained stable using the method of Agresti and Caffo’s adjusted Wald CIs that add “one success and one failure” to each group.22,23 We calculated the 95% CI for the difference between groups as a weighted combination of differences between the groups within the two randomisation stratifications. If the lower bound of the 95% CI for the difference was within the non-inferiority margin of 25%, we deemed eliglustat treatment non-inferior to imiglucerase. For the secondary objectives, if the lower bound of the exact 95% CI for the eliglustat group was more than 50%, we deemed eliglustat treatment successful at maintaining stability in most patients, irrespective of whether we deemed eliglustat non-inferior to imiglucerase. We analysed the secondary efficacy endpoints using ANCOVA. We did all statistical analyses with SAS version 9 or higher. This trial is registered with ClinicalTrials.gov, number NCT00943111, and the European Union Drug Regulating Authorities Clinical Trials database, number 2008-005223-28.

This trial was funded by Genzyme, a Sanofi company. Genzyme developed the design and set-up of the trial in collaboration with study investigators and regulatory authorities. Research associates contracted by Genzyme monitored study data. Genzyme Biomedical Data Science and Informatics did the analyses. Genzyme provided funding for a medical writer to assist the authors. All authors had full access to the study data. The corresponding author takes responsibility for the integrity of the data and the accuracy of the data analysis, and all authors jointly decided to submit for publication.

Results Between Sept 15, 2009, and Nov 9, 2011, we screened 209 eligible patients. Of these, we randomly allocated 160 patients to oral eliglustat (106 patients [66%]) or imiglucerase infusions (54 patients [34%]). The per-protocol population consisted of 99 (93%) patients receiving eliglustat and 47 (87%) receiving imiglucerase (figure 1). Demographic and baseline data are shown in table 1. On average, study participants had been on ERT for about 10 years, with nearly 60% receiving doses of at least 35 U/kg www.thelancet.com Vol 385 June 13, 2015

Articles

www.thelancet.com Vol 385 June 13, 2015

Eliglustat Patients who met stability criteria (%)

100

95% 100%

Imiglucerase 96% 100% 93% 100%

96% 94%

94% 85%

80 60 40 20 0 Haemoglobin concentration

Platelet count

Spleen volume*

Liver volume

Composite

Figure 2: Percentage of patients who met stability criteria by individual components and composite endpoint 12 months after treatment initiation Per-protocol population. Error bars represent exact 95% CIs around the proportion. *Spleen percentages are based on the total number of non-splenectomised patients in each treatment group.

A

B Mean change from baseline (MN) Eliglustat Imiglucerase

4

–0·2 (–6·2%) –0·1 (–3·0%)

2

1·5 Mean liver volume (MN)

Mean spleen volume (MN)

6

0 Number of patients Eliglustat Imiglucerase

1·0

0·5

0·02 (1·8%) 0·03 (3·6%)

6

12

Baseline

6

12

70 39

70 39

70 39

98 47

97 47

98 47

C

D 300

140 120 Mean change from baseline (g/L) Eliglustat Imiglucerase

0

Baseline Number of patients Eliglustat Imiglucerase

Eliglustat Imiglucerase

Baseline

160

100

Mean change from baseline (MN)

0

3

6

–2·1 (–1·1%) 0·4 (0·4%) 9

12

Time after treatment started (months) 98 47

96 46

98 46

97 46

98 47

Mean platelet count (×109/L)

Mean haemoglobin concentration (g/L)

every 2 weeks. Genotype distribution was very similar in the two treatment groups, and most patients were classified as extensive metabolisers on the basis of CYP2D6 genotype. Most patients had at least one N370S allele and an intact spleen. Mean bone mineral density was in the normal range and maintained; mean bone marrow burden scores showed moderate infiltration of haemopoietic marrow and were also maintained (appendix). Demographics and baseline disease characteristics were very similar between the pre-ERT dose stratification groups in both trial arms (data not shown). Eliglustat met the criteria for non-inferiority to imiglucerase in maintaining stability of haematological and organ variables. Figure 2 shows that the composite primary endpoint was maintained after 12 months in 84 of 99 (85%) patients receiving eliglustat and 44 of 47 (94%) receiving imiglucerase, with a between-group difference of –8·8% (95% CI –17·6 to 4·2). The lower bound of the 95% CI for the difference in percentage (–17·6%) was within the prespecified threshold of –25%. Results were similar for the ITT population (88 patients [83%] of 106 on eliglustat vs 48 patients [91%] of 53 on imiglucerase, with a between-group difference of –7·5%; 95% CI –17·1 to 5·1; appendix). The percentage of patients who remained stable and the exact 95% CIs were also similar for the randomisation stratification groups within each trial arm (pretrial ERT dose of <35 U/kg every other week or ≥35 U/kg every other week). 12 months after the start of treatment, 92% of patients receiving eliglustat and 94% of those receiving imiglucerase had stable and normal haematological variables and organ volumes as defined by the secondary composite endpoint (appendix). Figure 2 also shows stability of the individual components of the composite primary endpoint. Differences between treatment groups in the percentage of patients maintaining stable individual variables were not significant. Figure 3 shows individual variables over time. The betweengroup least-square mean percentage changes from baseline in platelet count, liver volume, and spleen volume did not differ significantly (p>0·2 for all). We noted a small but significant difference in least-square mean absolute change for haemoglobin concentration favouring imiglucerase (p=0·025); however, the lower bound of the CI (–5·2 g/L) of this difference is not clinically significant (appendix). Of the 18 patients (15 given eliglustat and three given imiglucerase) who did not meet the primary endpoint, 16 missed on one criterion, one patient on eliglustat missed on two criteria (spleen volume and platelet count), and one patient on eliglustat was counted as a failure because this patient voluntarily switched to imiglucerase after 9 months, although all stability criteria were met at the time of the switch. Criteria that were not met in the eliglustat treatment group were haemoglobin concentration (four patients), platelet count (six patients), liver volume (three patients), and spleen volume (two patients). The criterion that was not met in the three patients treated with imiglucerase was liver volume.

250 200 150 100

Mean change from baseline (×109/L) Eliglustat Imiglucerase

50 0

Baseline

3

6

9·5 (3·8%) 6·0 (2·9%) 9

12

Time after treatment started (months) 98 47

96 46

98 46

97 46

98 47

Figure 3: Haematological and organ volume stability from baseline to 12 months Per-protocol population. (A) spleen volume, (B) liver volume, (C) haemoglobin concentration, and (D) platelet count. Error bars represent SDs.

Because we based the prespecified endpoint criteria on change from baseline without accounting for baseline values, patients could fail to reach the primary endpoint, but still achieve therapeutic goals recommended by Pastores and colleagues24 for patients receiving ERT, which occurred for 12 of 15 patients in the eliglustat group and all three patients given imiglucerase. By definition, the study population received long-term enzyme therapy before randomisation and thus had a mainly low baseline burden of disease. We noted no 2359

Articles

large differences between groups or changes during the study in quality-of-life assessments (appendix). In a treatment preference questionnaire at screening, 94% of patients in both per-protocol treatment groups (eliglustat 93 [94%] of 99; imiglucerase 44 [94%] of 47) expressed a preference for oral treatment when asked “Although you Eliglustat (n=106)

Imiglucerase (n=53)

Patients (%)

Events

Patients (%)

Events

97 (92%)

656

42 (79%)

141

Related*

40 (38%)

109

6 (11%)

10

Not related†

96 (91%)

547

41 (77%)

131

Mild

94 (89%)

478

37 (70%)

103

Moderate

49 (46%)

149

19 (36%)

32

Severe

13 (12%)

29

4 (8%)

6

Serious‡

11 (10%)

11

0

0

Any TEAE

Death

0

0

0

0

Infusion-associated reaction§

0

0

2 (4%)

6

TEAE leading to study drug discontinuation

2 (2%)

2

1 (2%)

1 ··

TEAEs in ≥5% of patients, irrespective of relation to treatment Arthralgia

16 (15%)

··

9 (17%)

Asthenia

9 (8%)

··

0

··

Back pain

13 (12%)

··

3 (6%)

··

Blood CPK concentration increased

7 (7%)

··

1 (2%)

··

Bone pain

6 (6%)

··

1 (2%)

··

Constipation

5 (5%)

··

0

··

Contusion

5 (5%)

··

0

··

Cough

7 (7%)

··

2 (4%)

·· ··

Diarrhoea

13 (12%)

··

2 (4%)

Dizziness

9 (8%)

··

0

··

Dyspepsia

7 (7%)

··

1 (2%)

··

Epistaxis

5 (5%)

··

0

··

Fatigue¶

15 (14%)

··

1 (2%)

··

Gastro-oesophageal reflux disease

7 (7%)

··

0

··

14 (13%)

··

1 (2%)

··

Hepatomegaly

1 (1%)

··

3 (6%)

··

Influenza

6 (6%)

··

2 (4%)

·· ··

Headache¶

Nasopharyngitis

11 (10%)

··

5 (9%)

Nausea¶

13 (12%)

··

0

··

Pain in extremity

12 (11%)

··

1 (2%)

··

Palpitations

5 (5%)

··

0

··

Rash

5 (5%)

··

0

··

11 (10%)

··

1 (2%)

··

Sinusitis Toothache

2 (2%)

··

3 (6%)

··

Upper abdominal pain¶

11 (10%)

··

0

··

Upper respiratory tract infection

11 (10%)

··

3 (6%)

··

Urinary tract infection

5 (5%)

··

5 (9%)

··

Viral gastroenteritis

5 (5%)

··

1 (2%)

··

TEAE=treatment-emergent adverse event. CPK=creatine phosphokinase. *Defined as possibly, probably, or definitely related. †Defined as not related or remote or unlikely. ‡Serious adverse events were appendicitis (n=1), diverticulitis (n=1), malignant hepatic neoplasm (n=1), uterine leiomyoma (n=1), syncope (n=2), myocardial infarction (n=1), ischaemic colitis (n=1), cholecystitis (n=1), joint dislocation (n=1), and mammoplasty (n=1). §Any adverse event related (ie, possible, probable, or definite) to and occurring during or just after imiglucerase infusion. ¶Indicates difference of more than 10% between patients given eliglustat and those given imiglucerase.

Table 2: Adverse events in the safety population

2360

have not received an oral treatment yet, if you could choose, which would you prefer: oral treatment or intravenous treatment?” At 12 months, all 93 patients given eliglustat who completed the survey confirmed their preference for oral treatment for the following reasons: “more convenient” (81%); “taken at home” (69%); “given by tablet” (59%); “felt better after treatment” (22%); and “causes fewer side-effects” (11%). Detailed responses to this survey are provided in the appendix. 3 months after treatment was started, plasma concentrations of the biomarkers glucosylceramide and GM3 decreased by more than 50% in the eliglustat group and were maintained within the healthy reference range (appendix), consistent with the mechanism of action as an inhibitor of glucosylceramide synthase.7 Differences between groups at baseline and during the study were minimal for plasma chitotriosidase activity and ceramide, macrophage inflammatory protein 1β, and sphingomyelin concentrations (appendix). Most adverse events were non-serious and mild or moderate in severity (table 2). 11 serious adverse events occurred in 11 patients in the eliglustat group (10%), none considered related to treatment; several were possibly related to underlying Gaucher’s disease, including hepatocellular carcinoma (in retrospect, present at baseline),25 cholecystitis,26 and joint dislocation,5 and others were hospital admissions for intercurrent events (diverticulitis, uterine leiomyoma, appendicitis, cholecystitis, mammoplasty, and ischaemic colitis). Two (2%) patients given eliglustat and one (2%) given imiglucerase withdrew because of the following adverse events: palpitations without clinically relevant ECG findings, deemed possibly treatment-related (patient given eliglustat, day 198); myocardial infarction, deemed unrelated to treatment (patient given eliglustat, day 237); and psychotic disorder, deemed unrelated to treatment (patient given imiglucerase, day 172). The most common adverse events deemed related to eliglustat were diarrhoea (five [5%] patients), arthralgia (four [4%] patients), fatigue (four [4%] patients), and headache (four [4%] patients). Overall, most episodes of diarrhoea were deemed unrelated to treatment; five related events were mild, one was moderate, and none needed a treatment or dose change. ECG analysis showed no significant effect of eliglustat on heart rate or cardiac repolarisation. No patient reached the predefined threshold for QTcF abnormality (>480 ms or >60 ms increase from baseline). Four (4%) patients in the eliglustat group reached the threshold for PR abnormality (>200 ms and increase from baseline of ≥25%). The effect of eliglustat on PR interval was less than that seen with drugs known to prolong atrioventricular conduction (eg, β blockers). Eliglustat had a 2–3 ms (upper bound of 90% CI 4·8 ms) effect on cardiac depolarisation (QRS duration), which was not time-dependent or dose-dependent, and only slightly greater than that of imiglucerase (maximum 2·5 ms, upper bound of 90% CI 4·1 ms). One (1%) eliglustat patient reached the threshold www.thelancet.com Vol 385 June 13, 2015

Articles

for predefined QRS abnormality criteria (QRS ≥120 ms). Of four syncopes in three patients, all were vasovagal with predisposing risk factors (blood draw, fasting conditions, and pain); one patient had a history of syncope. All events resolved and resulted in neither dose reductions nor trial withdrawal. ECGs after events showed no arrhythmia.

Discussion Eliglustat, an investigational oral SRT, was non-inferior to imiglucerase in maintaining stability of haematological characteristics and organ volumes in adults with Gaucher’s disease, according to both the primary and secondary endpoints. Disease in cortical and marrow compartments of bone and quality-of-life variables also remained stable in both groups of patients receiving eliglustat and imiglucerase. These findings extend the efficacy profile of eliglustat beyond treatment-naive adult patients with Gaucher’s disease type 112–14 to include maintenance treatment in adults who have been stabilised on enzyme therapy (panel). Eliglustat was well tolerated in this study. Adverse events noted in patients given eliglustat were mainly mild to moderate in severity. Serious adverse events were related to hospital admission for intercurrent illnesses or, possibly, to underlying Gaucher’s disease. Eliglustat had no significant effects on heart rate or cardiac repolarisation. Mild or moderate diarrhoea deemed related to treatment was reported in five (5%) patients given eliglustat and did not need treatment or dose changes. The fact that we noted more adverse events in the eliglustat than the imiglucerase group was expected because adverse events reported with imiglucerase typically occur within the first year of treatment,31 and patients in this study had been taking imiglucerase for at least 3 years before enrolment. We did not see the frequency of some adverse events reported with miglustat (>80% diarrhoea, about 55–65% weight loss, and about 37% tremor)8–10,32 with eliglustat. These adverse events seem to represent off-target effects of miglustat, an N-butyliminosugar that is a non-specific glucosidase inhibitor known to inhibit intestinal disaccharidases and cross the blood–brain barrier.33 Eliglustat is a potent and specific inhibitor of glucosylceramide synthase, with no measurable inhibition of glycosidases or intestinal disaccharidases, and does not cross the blood–brain barrier.7 An open-label, non-inferiority design has limitations. Although a double-blind design is preferable, this design would have needed participants to take an oral drug (placebo or eliglustat) daily and receive intravenous placebo or imiglucerase infusions every 2 weeks. Although patients were not masked to treatment allocation, all four components of the composite efficacy endpoint are objective measurements not prone to bias, and all imaging, ECG, and Holter monitoring, and nerve conduction data were examined by external central readers masked to treatment assignments. Furthermore, the composite primary endpoint was a sensitive endpoint for a multisystem disease. The differences noted between www.thelancet.com Vol 385 June 13, 2015

Panel: Research in context Systematic review We searched PubMed for randomised clinical trials published up to Sept 30, 2014, in English. We used the search terms “Gaucher disease type 1”, “substrate reduction therapy”, “miglustat”, and “eliglustat”. Because eliglustat was an investigational product at the time of manuscript submission, the published data on eliglustat consisted of four publications from a phase 2 clinical trial in 26 treatment-naive patients with Gaucher’s disease type 1.11–14 Authors of these clinical publications reported significant reductions in spleen and liver volumes, and significant improvements in haemoglobin concentration, platelet count, and bone mineral density that continued or were maintained for up to 4 years after treatment initiation; one patient discontinued because of an adverse event that was deemed related to eliglustat. We refined our search by focusing on studies in which patients were switched from enzyme therapy to a substrate reduction therapy. We identified one randomised trial of miglustat in which 36 patients with Gaucher’s disease type 1 stabilised on enzyme therapy were randomly allocated miglustat, imiglucerase, or both.27 Because of equivocal treatment effects, short trial duration, and small number of patients, an additional clinical trial was requested by the European Medicines Agency.8 The subsequent non-inferiority trial,8 a 2 year, open-label study of 42 patients with Gaucher’s disease type 1 who had been stabilised after at least 3 years of previous enzyme therapy, met its primary efficacy endpoint of a less than 10% increase in liver volume; however, 21 patients (50%) discontinued treatment, with 13 of these due to adverse events (mostly gastrointestinal) and five due to worsening disease. Three small, open-label, non-comparative studies have also been done28–30 in which some or all patients switched from enzyme therapy to miglustat; these generally showed stable organ volumes and haematological variables after switching to miglustat. Because of its less favourable efficacy, safety, and tolerability profile than enzyme therapy, miglustat is licensed only for patients who are not suitable candidates for enzyme therapy.8–10 Interpretation To our knowledge, the ENCORE trial is the first study to show that adults with Gaucher’s disease type 1 who are stable on intravenous infusions of enzyme therapy can safely maintain stable haematological and organ measures after switching to oral eliglustat treatment.

treatments were within the non-inferiority margin and not clinically meaningful because most patients remained stable within individual component thresholds. Treatment with eliglustat and imiglucerase had similar effects on biomarkers of Gaucher’s disease, apart from decreased plasma glucosylceramide and GM3 concentrations with eliglustat, consistent with its mechanism of action. A diffusible small molecule that safely modulates synthesis of glucosylceramide and generation of cognate sphingolipids at the level of glucosylceramide formation could have more fundamental effects in Gaucher’s disease than enzyme augmentation treatments, which selectively target macrophages.34 With this fact in mind, eliglustat could have the potential to address skeletal,4,11 pulmonary,7 and other systemic manifestations of Gaucher’s disease refractory to existing treatment.3 In this phase 3 study, treatment with eliglustat maintained haematological and organ volume stability in patients with Gaucher’s disease type 1 who had reached therapeutic goals with ERT. Oral eliglustat is a promising first-line therapeutic alternative to imiglucerase as maintenance treatment in adult patients with this disease. 2361

Articles

Contributors TMC, GD, RC, MB, TAB, AMM, EL, and BR recruited patients and did the study research. TMC participated in writing of the report. GD, RC, MB, TAB, AMM, EL, BR, LR, JA, and ACP reviewed and edited initial and final versions. LR analysed safety data, JA did biostatistical analyses, and ACP designed the study. All authors reviewed and approved the submitted version. Declaration of interests TMC, GD, RC, MB, TAB, AMM, EL, and BR are principal investigators in the eliglustat ENCORE trial and have received honoraria from Genzyme, a Sanofi company. TMC, MB, EL, and BR have received travel reimbursement from Genzyme, a Sanofi company. TMC has been on the advisory board of Genzyme, a Sanofi company. MB, AMM, and EL are principal investigators in the eliglustat ENGAGE trial. MB is a member of the North American advisory board for the International Collaborative Gaucher Group Registry. TAB has received honoraria from Biomarin. EL has received honoraria and travel reimbursement from Shire. LR, ACP, and JA are employees of Genzyme, a Sanofi company. Acknowledgments This work was funded by Genzyme, a Sanofi company. The authors thank the patients and health-care professionals (appendix) who participated; Laurie LaRusso for medical writing support funded by Genzyme, a Sanofi company; Raymond Mankoski and Lisa Underhill, Genzyme Global Medical Affairs, and M Judith Peterschmitt and Gerald F Cox, Genzyme Clinical Development, for critical review. Work at the University of Cambridge was supported by grant MR/K015338/1 from the UK Medical Research Council and at Mount Sinai Hospital by grant UL1TR000067 from the National Center for Advancing Translational Sciences, National Institutes of Health. References 1 Grabowski G, Kolodny EH, Weinreb N. Gaucher disease: phenotypic and genetic variation. In: Scriver CR, Beaudet AL, Valle D, et al, eds. The online metabolic and molecular basis of inherited metabolic disease. New York: McGraw-Hill Companies, 2006. 2 Chen M, Wang J. Gaucher disease: review of the literature. Arch Pathol Lab Med 2008; 132: 851–53. 3 Grabowski GA. Phenotype, diagnosis, and treatment of Gaucher’s disease. Lancet 2008; 372: 1263–71. 4 Mistry PK, Deegan P, Vellodi A, Cole JA, Yeh M, Weinreb NJ. Timing of initiation of enzyme replacement therapy after diagnosis of type 1 Gaucher disease: effect on incidence of avascular necrosis. Br J Haematol 2009; 147: 561–70. 5 Deegan PB, Pavlova E, Tindall J, et al. Osseous manifestations of adult Gaucher disease in the era of enzyme replacement therapy. Medicine (Baltimore) 2011; 90: 52–60. 6 Kaplan P, Baris H, De Meirleir L, et al. Revised recommendations for the management of Gaucher disease in children. Eur J Pediatr 2013; 172: 447–58. 7 McEachern KA, Fung J, Komarnitsky S, et al. A specific and potent inhibitor of glucosylceramide synthase for substrate inhibition therapy of Gaucher disease. Mol Genet Metab 2007; 91: 259–67. 8 Cox TM, Amato D, Hollak CE, et al. Evaluation of miglustat as maintenance therapy after enzyme therapy in adults with stable type 1 Gaucher disease: a prospective, open-label non-inferiority study. Orphanet J Rare Dis 2012; 7: 102. 9 Zavesca package insert, Feb, 2008. http://www.accessdata.fda.gov/ drugsatfda_docs/label/2008/021348s005lbl.pdf (accessed April 23, 2014). 10 Zavesca summary of product characteristics, 2013. http://www.ema. europa.eu/docs/en_GB/document_library/EPAR_-_Product_ Information/human/000435/WC500046726.pdf (accessed April 23, 2014). 11 Kamath RS, Lukina E, Watman N, et al. Skeletal improvement in patients with Gaucher disease type 1: a phase 2 trial of oral eliglustat. Skeletal Radiol 2014; 43: 1353–60. 12 Lukina E, Watman N, Arreguin EA, et al. A phase 2 study of eliglustat tartrate (Genz-112638), an oral substrate reduction therapy for Gaucher disease type 1. Blood 2010; 116: 893–99. 13 Lukina E, Watman N, Arreguin EA, et al. Improvement in hematological, visceral, and skeletal manifestations of Gaucher disease type 1 with oral eliglustat tartrate (Genz-112638) treatment: 2-year results of a phase 2 study. Blood 2010; 116: 4095–98.

2362

14

15

16

17

18

19

20 21

22

23

24

25

26

27

28

29

30

31

32

33

34

Lukina E, Watman N, Dragosky M, et al. Eliglustat, an investigational oral therapy for Gaucher disease type 1: phase 2 trial results after 4 years of treatment. Blood Cells Mol Dis 2014; 53: 274–76. Mistry PK, Lukina E, Ben Turkia H, et al. Effect of oral eliglustat on splenomegaly in patients with Gaucher disease type 1: the ENGAGE randomized clinical trial. JAMA 2015; 313: 695–706. Barton NW, Brady RO, Dambrosia JM, et al. Replacement therapy for inherited enzyme deficiency—macrophage-targeted glucocerebrosidase for Gaucher’s disease. N Engl J Med 1991; 324: 1464–70. Kishnani PS, DiRocco M, Kaplan P, et al. A randomized trial comparing the efficacy and safety of imiglucerase (Cerezyme) infusions every 4 weeks versus every 2 weeks in the maintenance therapy of adult patients with Gaucher disease type 1. Mol Genet Metab 2009; 96: 164–70. Maas M, Hangartner T, Mariani G, et al. Recommendations for the assessment and monitoring of skeletal manifestations in children with Gaucher disease. Skeletal Radiol 2008; 37: 185–88. Robertson PL, Maas M, Goldblatt J. Semiquantitative assessment of skeletal response to enzyme replacement therapy for Gaucher’s disease using the bone marrow burden score. AJR Am J Roentgenol 2007; 188: 1521–28. Haddley K. Taliglucerase alfa for the treatment of Gaucher’s disease. Drugs Today (Barc) 2012; 48: 525–32. Hollak CE. An evidence-based review of the potential benefits of taliglucerase alfa in the treatment of patients with Gaucher disease. Core Evid 2012; 7: 15–20. Agresti A, Caffo B. Simple and effective confidence intervals for proportions and difference of proportions results from adding two successes and two failures. Am Stat 2000; 54: 280–88. Dann RS, Koch GG. Methods for one-sided testing of the difference between proportions and sample size considerations related to non-inferiority clinical trials. Pharm Stat 2008; 7: 130–41. Pastores GM, Weinreb NJ, Aerts H, et al. Therapeutic goals in the treatment of Gaucher disease. Semin Hematol 2004; 41 (4 suppl 5): 4–14. Weinreb NJ, Lee RE. Causes of death due to hematological and non-hematological cancers in 57 US patients with type 1 Gaucher Disease who were never treated with enzyme replacement therapy. Crit Rev Oncog 2013; 18: 177–95. Taddei TH, Dziura J, Chen S, et al. High incidence of cholesterol gallstone disease in type 1 Gaucher disease: characterizing the biliary phenotype of type 1 Gaucher disease. J Inherit Metab Dis 2010; 33: 291–300. Elstein D, Dweck A, Attias D, et al. Oral maintenance clinical trial with miglustat for type I Gaucher disease: switch from or combination with intravenous enzyme replacement. Blood 2007; 110: 2296–301. Giraldo P, Alfonso P, Atutxa K, et al. Real-world clinical experience with long-term miglustat maintenance therapy in type 1 Gaucher disease: the ZAGAL project. Haematologica 2009; 94: 1771–75. Giraldo P, Latre P, Alfonso P, et al. Short-term effect of miglustat in every day clinical use in treatment-naive or previously treated patients with type 1 Gaucher’s disease. Haematologica 2006; 91: 703–06. Kuter DL, Mehta A, Hollak CE, et al. Miglustat therapy in type 1 Gaucher disease: clinical and safety outcomes in a multicenter retrospective cohort study. Blood Cells Mol Dis 2013; 51: 116–24. Cerezyme summary of product characteristics, 2009. http://www. ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_ Information/human/000157/WC500024112.pdf (accessed April 23, 2014). Cox T, Lachmann R, Hollak C, et al. Novel oral treatment of Gaucher’s disease with N-butyldeoxynojirimycin (OGT 918) to decrease substrate biosynthesis. Lancet 2000; 355: 1481–85. Ross DJ, Spira S, Buchbinder NA. Gaucher cells in pulmonarycapillary blood in association with pulmonary hypertension. N Engl J Med 1997; 336: 379–81. Mistry PK, Wraight EP, Cox TM. Therapeutic delivery of proteins to macrophages: implications for treatment of Gaucher’s disease. Lancet 1996; 348: 1555–59.

www.thelancet.com Vol 385 June 13, 2015