Long‐Term Safety and Efficacy of Mepolizumab in Eosinophilic Granulomatosis With Polyangiitis

Abstract

Objective

Eosinophilic granulomatosis with polyangiitis (EGPA) is a rare, relapsing, inflammatory disease. Management of EGPA predominantly relies on oral corticosteroids (OCS), which are associated with many adverse effects. The phase 3 MIRRA trial demonstrated efficacy and safety of mepolizumab, anti‐interleukin‐5 biologic, for EGPA. This open‐label extension (OLE) of MIRRA assessed long‐term safety and OCS‐sparing effects of mepolizumab.

Methods

The OLE (NCT03298061) was a multicenter study that enrolled patients from MIRRA who required OCS ≥5 mg/day up to six months after the end of MIRRA. All patients received mepolizumab 300 mg subcutaneously every four weeks plus standard of care until mepolizumab was discontinued or became approved and reimbursed for EGPA in the respective country. Key outcomes included adverse events (AEs) and use of OCS.

Results

One hundred patients were enrolled in the OLE. Mean (SD) and median (min–max) exposure during OLE was 38.5 (27.0) and 27.0 (1.0–89.0) months. On‐treatment AEs were experienced by 98% of patients (43% treatment related; most frequent: injection site reaction [10%]) and serious AEs by 38% of patients (6% treatment related) with no new safety signals versus MIRRA identified. Median (Q1–Q3) OCS dose decreased from 10.0 (7.8–15.0) mg/day at OLE baseline to 5.0 (0.0–10.0) mg/day at study exit. Proportion of patients using OCS >7.5 mg/day decreased from 75% at baseline to 32% at study exit; 28% of patients discontinued OCS.

Conclusion

Long‐term use of mepolizumab to treat EGPA was well tolerated and resulted in sustained reductions in OCS use.

INTRODUCTION

Eosinophilic granulomatosis with polyangiitis (EGPA) is a rare, progressive, relapsing/remitting inflammatory disease in which elevated eosinophil counts, asthma, sino‐nasal disease, pulmonary infiltrates, other organ manifestations, and vasculitis may cause significant morbidity. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ EGPA treatment has historically relied on oral corticosteroids (OCS), other immunosuppressive drugs, and cytotoxic therapies, although more recently biologic treatments have become available to patients. ¹ , ⁹ , ¹⁰ Achieving remission (controlled symptoms in absence of clinical signs of disease) with low‐dose OCS alone is often associated with relapse. ¹ As prolonged use of OCS increases the risk of adverse effects over time and is associated with acute infections, there is a pressing need for effective treatment strategies that spare OCS. ¹¹ , ¹² , ¹³

Mepolizumab is a humanized monoclonal antibody that specifically targets interleukin (IL)‐5 to reduce proliferation, activation, and survival of eosinophils, and is approved in multiple regions worldwide for the treatment of severe asthma with an eosinophilic phenotype, chronic rhinosinusitis with nasal polyps, hypereosinophilic syndrome (HES), and EGPA. ¹⁴ , ¹⁵ , ¹⁶ Mepolizumab is recommended for active nonsevere EGPA as well as for add‐on therapy to maintain remission in severe or nonsevere EGPA. ¹ , ⁹ , ¹⁰ The phase 3 placebo‐controlled MIRRA trial (NCT02020889) demonstrated that patients with nonsevere relapsing or refractory EGPA undergoing IL‐5 inhibition through treatment with mepolizumab achieved higher rates of remission, including at both weeks 36 and 48 of treatment, and had significantly fewer relapses over 52 weeks of treatment compared with those patients receiving placebo. ¹⁷ Additionally, MIRRA demonstrated OCS‐sparing benefits patients treated with mepolizumab had less cumulative exposure to OCS over 52 weeks (4.3 mg/day lower for mepolizumab) as compared with patients receiving placebo. ¹⁸ The efficacy of both mepolizumab and the anti‐IL‐5 receptor monoclonal antibody benralizumab were further reinforced in the recent noninferiority phase 3 trial of benralizumab or mepolizumab for relapsing or refractory EGPA. ¹⁹

The safety and clinical benefits of receiving mepolizumab for up to one year have been demonstrated, but systematic reports of long‐term use in EGPA are limited. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ A real‐world, observational, multicenter European study demonstrated the effectiveness of mepolizumab in reducing disease activity and use of OCS up to 24 months, ²² and results of a real‐world observational study in Japan (MARS study) demonstrated the safety, sustained OCS‐sparing benefits and long‐term disease control after a minimum of 192 weeks of treatment. ²³ This open‐label extension (OLE) study aimed to provide continued mepolizumab access while monitoring the long‐term extended safety profile and OCS‐sparing benefits of mepolizumab to patients with EGPA who participated in MIRRA until it received regulatory approval and became available in each respective country.

PATIENTS AND METHODS

Study design and treatments

The OLE of MIRRA (GSK ID: MEA116841, NCT03298061) was a phase 3, multicenter, OLE study conducted at 31 sites in seven countries. Between April 14, 2015, and February 16, 2023 (latest completion date), the OLE enabled eligible patients from MIRRA to receive open‐label mepolizumab before local market approval and reimbursement of mepolizumab was granted (or until July 2022 in the United Kingdom, as requested by the UK regulatory authority). For patients who received mepolizumab during MIRRA, this was a reintroduction of mepolizumab after being off treatment for at least eight weeks for the posttreatment follow‐up period. All patients received mepolizumab 300 mg subcutaneously every four weeks, in addition to standard of care. Mepolizumab treatment was stopped if there was a lack of benefit as determined by the physician (if a treatment‐limiting adverse event [AE] occurred or because of pregnancy) (Figure 1A). There was no guidance on tapering of OCS during the OLE and doses were reduced at investigator discretion.

Figure 1.

(A) OLE study design. (B) Patient disposition. *Patients had either completed the MIRRA trial and reached the week 60 timepoint or were withdrawn prematurely from MIRRA and had reached the date that would have been week 60 if they had completed the study. ^†Eligible patients could initiate mepolizumab under the OLE within a six‐month period starting from week 60 (or what would have been week 60) of MIRRA. Initiation of mepolizumab treatment under the OLE more than six months after the MIRRA week 60 timepoint was considered on an individual basis by the GSK Medical Monitor. ^‡Any patient discontinuing mepolizumab during the OLE was assessed 12 weeks after receiving their last mepolizumab dose at a follow‐up or end of program visit. If the patient was lost to follow‐up, an Early Withdrawal form was completed. ^§Patients marked as “completed treatment” remained in the study and received mepolizumab until the study end at their local site. ^¶Study end in the United Kingdom was not determined by approval. UK eligible patients could continue to receive mepolizumab under the OLE until July 2022. AE, adverse event; EGPA, eosinophilic granulomatosis with polyangiitis; GSK, GlaxoSmithKline; OLE, open‐label extension; SC, subcutaneous; SoC, standard of care.

Patients

Patients who previously participated in MIRRA in Belgium, Canada, France, Germany, Japan, the United Kingdom, and the United States were eligible for the OLE if they required ≥5 mg/day of prednisolone (or equivalent) for the control of their EGPA within six months after completing participation in MIRRA. Eligible patients included those who withdrew prematurely and reached what would have been the end of study timepoint. MIRRA inclusion and exclusion criteria have been published; key MIRRA criteria and full OLE eligibility criteria are shown in Supplementary Table 1. ¹⁷ Full details on how patient race and ethnicity were determined are shown in the Supplementary Material. Briefly, MIRRA included patients ≥18 years of age diagnosed with relapsing/refractory EGPA receiving a stable dose of OCS ≥7.5 mg/day for at least four weeks before baseline. Patients were excluded from MIRRA if they had life‐threatening or organ‐threatening EGPA or were receiving a dose of OCS >50 mg/day. ¹⁷ The 1990 American College of Rheumatology (ACR) classification criteria for EGPA were modified for the purposes for MIRRA to ensure recruited patients had a diagnosis of EGPA including key manifestations of the disease (asthma [remission or active] plus eosinophilia and two other common features of EGPA). ²⁴

Outcomes and assessments

The OLE was designed to provide mepolizumab to eligible patients who had participated in MIRRA while monitoring long‐term safety; formal endpoints were not defined. Patients who received at least one dose of open‐label mepolizumab defined the safety population for which all outcomes are reported. Outcomes are also described by subgroups according to whether patients previously received mepolizumab or placebo during MIRRA. Outcomes include patient demographic characteristics at OLE baseline, treatment completion rate (defined as continuing mepolizumab treatment until commercial availability), mepolizumab exposure duration, number of treatments administered, and patient years of exposure. Safety‐related assessments included AEs, serious AEs (SAEs), and AEs of special interest (AESI) throughout the study, collected through an electronic clinical report form.

On‐treatment AEs were defined as any untoward medical occurrence occurring from administration of the first treatment up to 28 days after the last study treatment. SAEs were defined as AEs that resulted in death, were life threatening, required hospitalization or prolongation of existing hospitalization, or resulted in disability or incapacity, a congenital anomaly or birth defect, liver injury and impaired liver function, or other serious events in the investigator's medical or scientific judgment. AEs and SAEs were classified as treatment related or not related to study treatment based on the opinion of the investigator. AESI included systemic reactions, including allergic or hypersensitivity and nonallergic reactions, local injection site reactions, serious and opportunistic infections, neoplasms, malignancies, cardiac disorders, including serious cardiac, vascular or thromboembolic events and ischemic events.

Post hoc outcomes included change in dose of OCS over the course of the study. Data on daily dose of OCS were gathered until a patient discontinued mepolizumab (last dose +28 days), study withdrawal, or completion, whichever occurred earlier. OCS doses are reported at MIRRA baseline, OLE baseline, and at three‐month periods while the patient was receiving treatment. The final three‐month period of available OCS data up until the patient either reached study end or discontinued treatment early defined the study exit timepoint. All doses of OCS are reported as prednisolone equivalent.

Sample size and statistical analysis

No formal significance testing was conducted, and sample size considerations were not applicable. The population sample consisted of the number of patients who were eligible from MIRRA and desired access to mepolizumab. Exposure time was reported as total patient‐years exposure, mean (SD), and median (min–max) months of exposure during the OLE. Safety‐related outcomes were summarized as the proportion of patients and exposure‐adjusted rate of events per 1,000 patient‐years (total number of AEs/[total duration of exposure in days/365.25]×1,000) to account for the length of exposure observed within this study.

Dose of OCS was summarized (post hoc) as mean (SD) and median (lower quartile [Q1], upper quartile [Q3]) dose, and as the number of patients on doses of 0, ≤4, ≤7.5, or >7.5 mg/day by three‐monthly periods while on treatment and at study exit, and as the percent change from MIRRA baseline at study timepoints and study exit. The threshold of OCS dose ≤7.5 mg/day was based on the OCS requirement for the EULAR definition of remission, whereas the dose of ≤4 mg formed part of the more stringent remission criteria used in MIRRA. ¹⁰ , ¹⁷

Ethics and approval

The study was conducted in accordance with International Council for Harmonisation (ICH) Good Clinical Practice, applicable patient privacy requirements, the Declaration of Helsinki, and according to the protocol. Written informed consent was obtained from each patient before performing any study‐specific procedures and the study was monitored according to ICH E6, Section 5.18. The local institutional review board or ethics committee at each site oversaw trial procedures (see Supplementary Material for further details).

RESULTS

Patient population

Of the 136 patients enrolled in MIRRA, 119 were in countries eligible for the OLE; patients from Italy and Spain received mepolizumab in a separate extension study. In the OLE, 101 patients were screened and 100 received treatment (Figure 1B). The mean (SD) age at OLE baseline was 49.6 (13.9) years, 57% of the patients were female, and 91% of patients were White. Patients were treated in seven countries with the largest proportion in the United States (49%), followed by Germany (15%), the United Kingdom, and France (11% each) (Table 1). Patients from the United States, Japan, and Canada were the first to reach the end of the OLE as the first countries to receive regulatory approval and local reimbursement for mepolizumab in EGPA in 2018 (Figure 1A).

Table 1.

Demographic characteristics at OLE baseline, location of study population, and clinical features at MIRRA baseline^*

	Total population, mepolizumab 300 mg SC (n = 100)	Previous placebo in MIRRA (n = 51)	Previous mepolizumab in MIRRA (n = 49)
At OLE baseline
Female, n (%)	57 (57)	29 (57)	28 (57)
Age (years)
Mean (SD)	49.6 (13.9)	49.7 (15.0)	49.4 (12.9)
19–64, n (%)	87 (87)	44 (86)	43 (88)
≥65, n (%)	13 (13)	7 (14)	6 (12)
Ethnicity, n (%)
Hispanic or Latino	1 (1)	0	1 (2)
Not Hispanic or Latino	99 (99)	51 (100)	48 (98)
Race, n (%)
American Indian or Alaskan Native	1 (1)	0	1 (2)
Asian	8 (8)	5 (10)	3 (6)
White	91 (91)	46 (90)	45 (92)
Country, n (%)
Belgium	3 (3)	0	3 (6)
Canada	5 (5)	2 (4)	3 (6)
France	11 (11)	6 (12)	5 (10)
Germany	15 (15)	10 (20)	5 (10)
Japan	6 (6)	3 (6)	3 (6)
United Kingdom	11 (11)	6 (12)	5 (10)
United States	49 (49)	24 (47)	25 (51)
At MIRRA baseline
BVAS total score
Mean (SD)	3.1 (4.61)	3.2 (4.12)	3.1 (5.12)
Median (Q1–Q3)	1.0 (0.0–4.0)	1.0 (0.0–5.0)	1.0 (0.0–4.0)
Min–max	0–22	0–19	0–22
VDI score
Mean (SD)	4.3 (2.73)	3.9 (2.59)	4.7 (2.84)
Median (Q1–Q3)	4.0 (2.0–6.0)	4.0 (2.0–6.0)	4.0 (3.0–6.0)
Min–max	0–12	0–10	0–12
History of positive ANCA test, n (%)	15 (15)	8 (16)	7 (14)

^*ANCA, antineutrophil cytoplasmic antibody; BVAS, Birmingham Vasculitis Activity Score; OLE, open‐label extension; Q1, lower quartile; Q3, upper quartile; SC, subcutaneous; VDI, Vasculitis Damage Index.

Overall, 73 (73%) patients completed the study and 27 (27%) discontinued early (Figure 1B). Treatment was completed by 71 (71%) patients, who received mepolizumab until the end of the study in their location and subsequently could receive commercially available mepolizumab. Of the 27 patients who discontinued the study early, the primary reasons for study withdrawal were patient decision (n = 10), physician decision (n = 7), lack of efficacy (n = 6), AEs (n = 3), and protocol deviation (n = 1) (Figure 1B). Additional details of reasons for discontinuation are shown in Supplementary Table 2; in total across all primary and subreasons, lack of, or partial efficacy or benefit or wanting to try another biologic was identified as a contributing reason for study withdrawal for 10 patients.

Exposure

In the OLE, patients were exposed to mepolizumab for a mean (SD) of 38.5 (27.0) months (3.2 [2.25] years); the median length of exposure was 27.0 months with a minimum of 1 month and a maximum of 89 months (7.4 years) (Table 2). The total mepolizumab exposure in the OLE was 320.5 patient‐years. Exposure among patients who completed treatment versus those who discontinued early is shown in Supplementary Table 3. In addition, 49 patients had received previous mepolizumab treatment under MIRRA for a full year, with the exception of two patients who were withdrawn from MIRRA after 3.7 months and 9.1 months, respectively. For these 49 patients who previously received mepolizumab the time between the end of treatment with mepolizumab in MIRRA and enrollment in the OLE varied between 8 weeks (MIRRA follow‐up period) and 20 months. Treatment adherence was incomplete with 38 patients missing at least one mepolizumab dose (Table 2). Consecutive missed doses were numerically more common among patients who discontinued early with six (21%) patients who discontinued missing three consecutive doses versus two (3%) patients who completed the doses (Supplementary Table 3).

Table 2.

Mepolizumab completions, discontinuations, exposure, and missed doses^*

	Total population, mepolizumab 300 mg SC (n = 100)	Previous placebo in MIRRA (n = 51)	Previous mepolizumab in MIRRA (n = 49)
Treatments status, n (%)
Completed	71 (71)	38 (75)	33 (67)
Discontinued early	29 (29)	13 (25)	16 (33) a
Primary reason for treatment discontinuation, b n (%)
AE	3 (3)	2 (4)	1 (2)
Lack of efficacy	4 (4)	1 (2)	3 (6)
Protocol deviation	1 (1)	1 (2)	0
Physician decision	9 (9)	2 (4)	7 (14)
Withdrawal by patient	12 (12)	7 (14)	5 (10)
Treatments administered
Mean (SD)	39.5 (28.4)	39.4 (27.8)	39.6 (29.3)
Median (range)	29.0 (1–97)	30.0 (1–96)	28.0 (3–97)
Range of exposure, months c
Mean (SD)	38.5 (27.0)	38.6 (26.5)	38.3 (27.7)
Median (min–max)	27.0 (1–89)	28.0 (1–88)	26.0 (3–89)
Total patient‐years exposure d	320.5	164.1	156.4
Missed doses, n (%)
No missed doses	62 (62)	26 (51)	36 (73)
≥1 missed dose	38 (38)	25 (49)	13 (27)
≥2 missed doses	22 (22)	16 (31)	6 (12)
2 consecutive missed doses	13 (13)	9 (18)	4 (8)
≥3 missed doses	17 (17)	11 (22)	6 (12)
3 consecutive missed doses	8 (8)	5 (10)	3 (6)

^*AE, adverse event; SC, subcutaneous.

^a Two patients in the previous mepolizumab group completed the study but discontinued treatment early (see Supplementary Table 2).

^b Patients may have only one primary reason for treatment discontinuation, in five cases this differed from the primary reason given for study withdrawal (see Supplementary Table 2).

^c Exposure (therapeutic coverage) = treatment stop date – treatment start date + 29 days.

^d Sum across patients of (treatment stop date – treatment start date + 29 days)/365.25.

Safety

On‐treatment AEs were experienced by 98 patients (98%; exposure‐adjusted rate 4,708.0 events/1,000 patient‐years exposure), with 43 patients having treatment‐related AEs (43%; 508.5 events/1,000 patient‐years exposure) (Table 3; Supplementary Table 4). The most frequent AEs were nasopharyngitis (33%; 352.6 events/1,000 patient‐years), upper respiratory tract infections (31%; 134.2 events/1,000 patient‐years), worsening of asthma (30%; 293.3 events/1,000 patient‐years), sinusitis (30%; 162.2 events/1,000 patient‐years), and bronchitis (29%; 152.9 events/1,000 patient‐years) (Supplementary Table 4).

Table 3.

Summary of AEs, SAEs, and AESI^*

	Total population, mepolizumab 300 mg SC (n = 100)	Previous placebo in MIRRA (n = 51)	Previous mepolizumab in MIRRA (n = 49)
Number of patients with AEs, n (%)
On‐treatment AEs a	98 (98)	50 (98)	48 (98)
Posttreatment AEs a	18 (18)	6 (12)	12 (24)
Treatment‐related AEs	43 (43)	21 (41)	22 (45)
AEs leading to permanent discontinuation of study treatment or withdrawal from study	3 (3)	2 (4)	1 (2)
Number of patients with SAEs, a n (%)
On‐treatment SAEs a	38 (38)	16 (31)	22 (45)
Posttreatment SAEs a	3 (3)	2 (4)	1 (2)
Treatment‐related SAEs (nonfatal)	6 (6)	2 (4)	4 (8)
Fatal SAEs	1 (1)	1 (2)	0
Treatment‐related fatal SAEs	0	0	0
Number of patients with on‐treatment AESI, b n (%)
Systemic reactions	4 (4)	3 (6)	1 (2)
Hypersensitivity reactions	3 (3)	3 (6)	0
Nonallergic reactions	1 (1)	0	1 (2)
Anaphylaxis	0	0	0
Local injection site reactions	20 (20)	10 (20)	10 (20)
All infections c	86 (86)	42 (82)	44 (90)
Serious infections	18 (18)	8 (16)	10 (20)
Potential opportunistic infections d	3 (3)	2 (4)	1 (2)
Neoplasms benign, malignant, and unspecified (including cysts and polyps)	10 (10)	4 (8)	6 (12)
Malignancies e	4 (4)	1 (2)	3 (6)
Cardiac disorders f	12 (12)	5 (10)	7 (14)
Serious cardiac disorders	7 (7)	3 (6)	4 (8)
Serious cardiac, vascular, or thromboembolic events g	9 (9)	5 (10)	4 (8)
Serious ischemic events h	3 (3)	1 (2)	2 (4)
Number of patients with on‐treatment treatment‐related AESIs, n (%)
Systemic reactions	4 (4)	3 (6)	1 (2)
Hypersensitivity reactions	3 (3)	3 (6))	0 (0)
Nonallergic reactions	1 (1)	0 (0)	1 (1)
Anaphylaxis	0	0	0
Local injection site reactions	17 (17)	7 (14)	10 (20)
All infections c	15 (15)	9 (18)	6 (12)
Serious infections	4 (4)	2 (4)	2 (4)
Potential opportunistic infections d	1 (1)	1 (2)	0 (0)
Neoplasms benign, malignant, and unspecified (including cysts and polyps)	4 (4)	1 (2)	3 (6)
Malignancies e	2 (2)	0 (0)	2 (4)
Cardiac disorders f	0	0	0
Serious cardiac disorders	0	0	0
Serious cardiac, vascular or thromboembolic events g	0	0	0
Serious ischemic events h	0	0	0

^*AE, adverse event; AESI, adverse event of special interest; MedDRA, Medical Dictionary for Regulatory Activities; SAE, serious adverse event; SC, subcutaneous; SMQ, standardized MedDRA queries.

^a AEs and SAEs were defined as on treatment if occurring after the date of first treatment within the study and before 28 days after last study treatment. AEs and SAEs occurring more than 28 days after last treatment were defined as post treatment.

^b AESIs included systemic reactions, including allergic/hypersensitivity and nonallergic reactions, local injection site reactions, serious and opportunistic infections, neoplasms, malignancies, cardiac disorders, including serious, cardiac, vascular, thromboembolic, and ischemic events.

^c Infections and infestations system organ class.

^d Identified from SMQs or events with the preferred term of herpes zoster.

^e Identified from SMQs.

^f Cardiac disorders system organ class.

^g Serious cardiac, vascular, thromboembolic events identified from cardiac disorders system organ class, vascular disorders system organ class, and SMQs.

^h A subset of serious cardiac, vascular, thromboembolic events, identified from MedDRA SMQ.

AEs led to permanent treatment discontinuation in three patients. Of these, one patient previously treated with placebo in MIRRA had fatal AEs of severe cardiac arrest, dyspnea, and hypoxia 258 days after the first dose and 12 days after the last dose of mepolizumab; this was the only death recorded in the study and these events were reported as not treatment related. The second patient had moderate‐severity colon cancer reported 628 days after the first dose and 9 days after the last dose of mepolizumab; this AE was reported treatment related by the investigator. The third patient experienced an AE of asthma and AEs of arthralgia, myalgia, and paresthesia 10 days and 12 days, respectively, after their first and only dose of mepolizumab; none of the AEs were reported as serious or treatment related.

On‐treatment SAEs were reported in 38% of patients (299.5 events/1,000 patient‐years) and were considered treatment related in 6% of patients (Table 3). The most frequent SAEs were worsening of asthma (6%; 21.8 events/1,000 patient‐years), worsening of EGPA (3%; 9.4 events/1,000 patient‐years), and pneumonia (3%; 21.8 events/1,000 patient‐years) (Supplementary Table 4). On‐treatment AESIs included systemic reactions (n = 4 of 100; treatment related: n = 4 of 4), local injection site reactions (n = 20 of 100; treatment related: n = 17 of 20), potential opportunistic infections (n = 3 of 100; treatment related: n = 1 of 3), neoplasms (n = 10 of 100; treatment related: 4 of 10), malignancies (n = 4 of 100; treatment related: n = 2 of 4), cardiac disorders (n = 12 of 100; treatment related: 0 of 12), and serious cardiac, vascular, or thromboembolic events (n = 9 of 100; treatment related: n = 0 of 9) (Table 3). No event of anaphylaxis was reported. The four malignancies reported during the OLE included one event each of renal cell carcinoma and colon cancer (both reported as treatment related by the investigator) and two events of basal cell carcinoma (not reported as treatment related). A breakdown of cardiovascular events is shown in Supplementary Table 5; no cardiac events were reported as treatment related.

Long‐term OCS‐sparing effects

At entry into the OLE, 75% of patients were taking OCS >7.5 mg/day with a median (Q1–Q3) dose of OCS of 10 (7.8–15.0) mg/day versus 12.5 (10.0–20.0) mg/day at MIRRA baseline (Figure 2A and B). At OLE baseline 2% (2 of 100) and 25% (25 of 100) of patients required OCS ≤4 or ≤7.5 mg/day, respectively, and by months 16 to 18, 38% (33 of 88) and 68% (60 of 88) had achieved a dose of ≤4 and ≤7.5 mg/day, respectively, with 20% (18 of 88) being able to stop OCS entirely. At study exit (the latest 3‐month period before either study completion or early discontinuation), 43% (43 of 100) and 68% (68 of 100) of patients were receiving an average dose of OCS ≤4 and ≤7.5 mg/day, respectively, with 28% (28 of 100) discontinuing OCS entirely (Figure 2A). Thirteen patients had no reduction or an increase in the dose of OCS at study exit versus MIRRA baseline, whereas 66 (66%) had reduced OCS by ≥50% (Figure 2C; Supplementary Table 6).

Figure 2.

(A) Number of patients receiving dose of OCS (prednisolone equivalent) by category. (B) Median (Q1–Q3) dose of OCS. (C) Percentage reduction in dose of OCS from MIRRA baseline, at exit from the OLE study. *Between the end of the MIRRA treatment period and initiation of OLE, mepolizumab treatment was discontinued for between 8 weeks and 20 months. ^†At OLE entry, two patients were classified as receiving OCS 0 mg/day because of a gap in the reporting of concomitant medication data. Both patients were receiving OCS of ≥5 mg/day and met all necessary eligibility criteria for the OLE. ^‡Study exit is defined as the latest three‐month period before patients either reached study end/completed (coinciding with commercial license of mepolizumab for EGPA in the relevant country) or discontinued treatment early. EGPA, eosinophilic granulomatosis with polyangiitis; OLE, open‐label extension; OCS, oral corticosteroids; Q1, lower quartile; Q3, upper quartile.

The median (Q1–Q3) daily dose of OCS decreased steadily from 10.0 (7.8–15.0) mg/day at OLE baseline over the first 12 months of OLE to 5.3 (3.0–9.9) mg/day at months 10 to 12 (n = 89) and remained reduced from baseline for most patients over the course of the study, with a median 5.0 (0.0–10.0) mg/day at study exit (Figure 2B). Median and mean doses of OCS and discontinuation of OCS during the OLE were similar regardless of prior mepolizumab or placebo treatment (Supplementary Figure 1A and B; Supplementary Table 7).

Patterns of use of OCS at the individual patient level over the course of the OLE are represented as heat maps shown in Supplementary Figure 2 (absolute dose) and Supplementary Figure 3 (percentage change compared with MIRRA baseline). In both plots, the steady and sustained shift in color toward pale blue (Supplementary Figure 2) or yellow (Supplementary Figure 3) for many patients indicates that patients were able to successfully taper OCS. Overall, 31 (31%) patients discontinued OCS completely for one or more three‐month period during the study. Darker blue (Supplementary Figure 2) and red colors (Supplementary Figure 3) represent higher doses of OCS demonstrating that some patients were not able to taper or experienced a relapse requiring increased treatment. For most patients, increases in use of OCS did not correspond to discontinuation of mepolizumab.

DISCUSSION

This OLE study of treatment with mepolizumab for patients from MIRRA adds to the body of evidence on the long‐term safety and OCS‐sparing effects of mepolizumab, ²² , ²³ , ²⁵ , ²⁶ , ²⁷ and demonstrates a positive long‐term benefit:risk profile of mepolizumab for patients with EGPA. More than two‐thirds of patients continued to receive mepolizumab treatment under the OLE until mepolizumab became commercially available, suggesting that long‐term IL‐5 inhibition by mepolizumab was generally well tolerated and provided OCS‐sparing benefits over several years.

In total the OLE captured 320.5 patient‐years of mepolizumab exposure, nearly five times the patient‐years observed in the parent MIRRA study with an overall safety profile consistent with that seen in MIRRA, ¹⁷ and prior studies in HES, ²⁸ , ²⁹ and EGPA, ²³ , ²⁵ with no new safety concerns identified. Previous long‐term international data on the use of mepolizumab beyond one year were available only for the mepolizumab 100 mg dose for severe asthma treatment ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ and for a higher dose (mepolizumab 750 mg) with variable dosing intervals in patients with HES. ³⁶

Of the patients who withdrew from the OLE early, few cases reported a lack of efficacy (6) or AEs (3) as the primary reason for early discontinuation, consistent with MIRRA in which only 3% of 68 patients receiving mepolizumab discontinued because of AEs. ¹⁷ The more commonly observed AEs in the OLE were consistent with typical manifestations of EGPA observed in other studies, with worsening asthma being the most common SAE and nasopharyngitis, upper respiratory tract infection, sinusitis, arthralgia, and headache among the most common AEs. ³ , ⁴ , ⁵ , ⁷ , ¹⁷ , ¹⁹ , ²² Systemic reactions and local injection site reactions were similar to those observed in MIRRA, with no anaphylaxis reported. Four patients reported malignancies during the OLE (two reported as treatment related by the investigator), which was similar to that reported in the long‐term, real‐world MARS study in Japan (six AEs of malignant tumors). ²³

As higher cumulative OCS exposure is associated with increased risk of AEs, ¹³ , ³⁷ minimizing OCS use is an important treatment goal. ¹ , ³⁸ In the OLE study, the median daily dose of OCS was reduced from 10 mg/day at baseline to 5 mg/day at study exit. In MIRRA, 54% of mepolizumab‐treated patients were on a dose of OCS ≤7.5 mg/day at weeks 36 and 48. The OLE permitted even greater OCS reductions, with 68% of patients receiving ≤7.5 mg/day, 43% on OCS ≤4 mg/day, and 28% having discontinued OCS at study exit.

The long‐term OCS‐sparing effects of mepolizumab for patients with EGPA were demonstrated in real‐world studies, ²² , ²³ , ²⁶ , ²⁷ including the MARS study in which 36% of patients discontinued OCS after a minimum 192 weeks of mepolizumab treatment. ²³ It is likely that benefits of mepolizumab accrue over time, with 31% of patients being OCS free for three months or longer during the OLE, an increase from 25% of patients treated with mepolizumab in MIRRA who accrued any OCS‐free weeks. ¹⁸ At an individual level, many patients experienced durable reductions in OCS use lasting from months to years, a clinically meaningful outcome from long‐term mepolizumab use.

OCS‐tapering in the OLE was reflective of real‐world clinical practice. Efforts to reduce OCS at expert sites, before approval of mepolizumab by the US Food and Drug Administration, reflects strong motivation by both patients and clinicians. Notably, the OCS reductions occurred even in the context of imperfect adherence (one or more doses of mepolizumab missed) as would be expected in real‐world practice.

There are several limitations of the current study, including the fact this was an open‐label, nonrandomized, single‐arm study with no comparator group, and variable length of follow‐up, and there was no defined OCS‐tapering protocol. Data on concomitant medications beyond OCS such as immunosuppressive, cytotoxic, or inhaled or topical therapies (such as for asthma or sinus disease) were not collected, nor were data on blood eosinophil counts, sequelae such as relapses of EGPA, remission status, Birmingham vasculitis activity scores, or organ manifestations. AEs were collected via a prespecified strategy on standardized electronic case report forms but were not adjudicated. As an extension from MIRRA, there is the potential that this study reflects a population that is more likely to tolerate and respond well to mepolizumab; however, only two (3%) patients treated with mepolizumab in MIRRA experienced treatment‐limiting AEs, and most patients chose to continue in the OLE. ¹⁷ Finally, 38% of patients missed at least one dose of mepolizumab and 17% missed up to three doses, which may have led to an underestimation of the OCS‐sparing benefit of mepolizumab.

This OLE study, the first long‐term prospective, international study for patients with EGPA treated with mepolizumab 300 mg subcutaneously (SC), extends and expands on previous long‐term follow‐up data with mepolizumab 300 mg SC. Incorporating AEs, exposure data, and OCS reduction, these results support a positive long‐term safety profile as well as sustained efficacy of prolonged use of mepolizumab in the treatment of EGPA. Studies are needed to examine the ability of mepolizumab to help achieve reduction of other immunosuppressive drugs, prevent organ damage associated with long‐term OCS use, and allow patients to achieve complete remission.

AUTHOR CONTRIBUTIONS

All authors contributed to at least one of the following manuscript preparation roles: conceptualization AND/OR methodology, software, investigation, formal analysis, data curation, visualization, and validation AND drafting or reviewing/editing the final draft. As corresponding author, Dr Khoury confirms that all authors have provided the final approval of the version to be published, and takes responsibility for the affirmations regarding article submission (eg, not under consideration by another journal), the integrity of the data presented, and the statements regarding compliance with institutional review board/Declaration of Helsinki requirements.

ROLE OF THE STUDY SPONSOR

GSK was involved in study design and implementation, as well as data collection, analysis, interpretation, writing the study report and reviewing this manuscript. Writing support provided by Alice Rees, PhD, was funded by GSK. Publication of this article was not contingent upon approval by GSK.

Supporting information

Disclosure Form:

Appendix S1: Supporting Information

APPENDIX A. EGPA MEPOLIZUMAB OPEN‐LABEL EXTENSION STUDY GROUP MEMBERS

Members of the EGPA Mepolizumab Open‐Label Extension study group are as follows: Arnaud Bourdin, Hôpital Arnaud de Villeneuve, Montpellier, France; Pascal Chanez, Hôpital Nord, Marseille, France; Anoop Chauhan, Queen Alexandra Hospital, Portsmouth, United Kingdom; Vincent Cottin, Hôpital Louis Pradel, Bron, France; Gerald Gleich, University of Utah, Murray, Utah; Berhnard Hellmich, Medius Klinik Kirchheim, Kirchheim‐Teck, Germany; Tomonori Ishii, Tohoku University Hospital, Sendai‐city, Miyagi, Japan; David Jayne, Addenbrookes Hospital NHS Trust, Cambridge, United Kingdom; Jean‐Emmanuel Kahn, Hôpital Foch, Suresnes, France; Peter Kern, Klinikum Fulda‐MVZ Osthessen, Hessen, Germany; Nader Khalidi, St. Joseph's Healthcare Hamilton, Hamilton, Ontario, Canada; Ina Koetter, Rheumaklinik Bad Bramstedt and Universitaetsklinikum Schleswig‐Holstein, Schleswig‐Holstein, Germany; Carol Langford, Cleveland Clinic Foundation, Cleveland, Ohio; Amy Markezich; Overlake Medical Center, Bellevue, Washington; Paul Monach, Boston University School of Medicine, Boston, Massachusetts; Shamsa Naveed, Glenfield Hospital, Leicester, United Kingdom; Peter Oelzner, Universitaetsklinikum Jena, Jena, Thueringen, Germany; Kenneth Pinna, Southwest Allergy and Asthma Clinic, St. George, Utah; Xavier Puechal, Hôpital Cochin, Paris, France; Emory H. Robinette, Pulmonary Research of Abingdon, Abingdon, Virginia; Florence Roufosse, Hôpital Erasme, Bruxelles, Belgium; Reinhard Voll, Universitaetsklinikum Freiburg, Freiburg, Germany; Andrew Schafer, Weill Medical College of Cornell University/New York Presbyterian Hospital, New York City, New York; Kaharu Sumino, Washington University School of Medicine, St. Louis, Missouri; Masami Taniguchi, Sagamihara National Hospital, Kanagawa, Japan; Phillip Lieberman, Baptist Medical Group, Germantown, Tennessee.