Real-world outcomes after anti–IL-5/anti–IL-5Rα treatment for hypereosinophilic syndrome: Systematic literature review

Abstract

Background

Anti–IL-5/anti–IL-5 receptor alpha (anti–IL-5/anti–IL-5Rα) treatments have potential to improve outcomes in hypereosinophilic syndrome (HES), as demonstrated in clinical trials.

Objective

We sought to understand the real-world outcomes of anti–IL-5/anti–IL-5Rα treatments in patients with HES.

Methods

This systematic literature review searched English-language articles (November 4, 2015, to October 31, 2023) and conference abstracts (January 1, 2021, to October 31, 2023) from MEDLINE/Embase and select conferences for publications containing outcomes for patients with HES receiving anti–IL-5 (mepolizumab [approved], reslizumab) or anti–IL-5Rα (benralizumab) treatment in the real world. Meta-analyses and clinical trials were excluded. Outcomes included treatment response, blood eosinophil (bEOS) counts, and oral corticosteroid (OCS) receipt. No formal bias assessment was conducted.

Results

Five cohort studies (192 participants) and 55 case reports (70 participants) were identified. In cohort studies, median age was 44.0 to 61.8 years; 33.3% to 62.9% were female, and 45.5% to 100% had idiopathic HES. Mean (standard deviation) age in case reports (n = 68/70) was 38.5 (23.7) years; 40.0% (n = 28/70) were female, and 67.2% (n = 39/58) had idiopathic HES. Across 4 cohort studies, most patients had reduced or no HES flares after mepolizumab or benralizumab. In patients from case reports, 13 had HES flares (mepolizumab, n = 12/54; benralizumab, n = 1/13) after anti–IL-5/anti–IL-5Rα initiation. Mean bEOS counts were reduced by 53.0% to 100% across 3 cohort studies and by 95.0% in case reports. Of those receiving benralizumab or mepolizumab, OCS dose was reduced in 49.0% to 100% of patients from cohort studies, and by ≥50% in 29 of 70 patients from case reports.

Conclusion

Real-world anti–IL-5/anti–IL-5Rα therapy in HES is associated with fewer flares as well as reductions in bEOS counts and OCS dose, although confirmatory studies in larger populations are needed.

Hypereosinophilic syndrome (HES) encompasses a group of rare hematologic disorders characterized by persistent hypereosinophilia, resulting in eosinophil-mediated organ damage/dysfunction.1, 2, 3, 4 HES is typically classified by distinct clinical, laboratory, and molecular features as follows: primary (neoplastic/clonal), secondary (reactive), and idiopathic (I-HES; without an identifiable nonhematologic secondary cause).5, 6, 7 The true prevalence of HES is unknown, although it is estimated that there are about 0.32 to 6.3 cases per 100,000 population.⁸

Clinical manifestations of HES are heterogenous and involve numerous organ systems including the skin, lungs, gastrointestinal tract, and nervous and cardiovascular systems.⁹ Diagnosis is thus complex and, in the absence of specific comprehensive guidelines, relies on excluding other eosinophilic conditions.^6,10

Without treatment, HES can be fatal; however, the 5-year survival rate with prompt management is >80%.¹¹ Treatment of HES relies on systemic therapies such as high-dose oral corticosteroids (OCS) as well as cytotoxic and/or immunosuppressive agents; however, patients may still experience suboptimal disease control with these approaches.^6,10, 11, 12, 13 In addition, long-term OCS receipt, even at low doses, is associated with multiple adverse effects, including osteoporosis, obesity, hypertension, diabetes, cataracts, and infections.^12,14, 15, 16, 17 Given the limitations of OCS, there was a need for additional options, and biologic agents have recently been developed that may aid in the treatment of certain HES subtypes.

Eosinophil-targeting treatments represent a therapeutic advance in the management of HES.^18,19 Mepolizumab is a humanized anti–IL-5 monoclonal antibody and is the only biologic approved for HES.^20,21 A phase 3 study demonstrated that the proportion of patients experiencing a HES flare was significantly lower with mepolizumab versus placebo (27.8% vs 55.6%, respectively; P = .002).²² Reslizumab is also a humanized anti–IL-5 monoclonal antibody with demonstrated efficacy in a phase 1/2 trial of 4 patients with refractory HES. Reslizumab was well tolerated and led to decreased eosinophil counts for up to 12 weeks; however, eosinophil counts increased as reslizumab levels waned, resulting in exacerbation of symptoms.23, 24, 25 Benralizumab is a humanized monoclonal antibody that binds specifically to the alpha subunit of the anti–IL-5 receptor (anti–IL-5Rα); it induces near-complete removal of eosinophils through enhanced antibody-dependent cellular cytotoxicity.²⁶ It is currently being investigated in the phase 3 double-blind, placebo-controlled NATRON trial (NCT04191304) in patients with HES.²⁷

Because the disease is so rare, there is limited understanding of the real-world use of anti–IL-5/anti–IL-5Rα treatments in patients with HES; hence, our objective was to perform a comprehensive systematic literature review (SLR) of observational cohort studies and case reports to further our understanding of the real-world use and associated outcomes of anti–IL-5/anti–IL-5Rα treatments for HES.

Methods

This SLR was designed, conducted, and reported according to the recommendations of the Cochrane Handbook of Systematic Reviews of Interventions²⁸ and the 2020 Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines.²⁹

Eligibility criteria

This SLR identified literature published from November 4, 2015 (full-length articles), or January 1, 2021 (conference abstracts), until October 31, 2023. Only full-text articles published in English or that had an English translation were included. Full eligibility criteria are listed in Table E1 in this article’s Online Repository available at www.jaci-global.org.

Patient population

Studies reporting on anti–IL-5/anti–IL-5Rα treatments in patients with HES were included. No patient age limit was applied.

Intervention and comparators

Studies were included if they contained details of benralizumab, mepolizumab, or reslizumab treatment after a HES diagnosis.

Data collection and outcome measures

Data were collected on study characteristics, patient demographics/characteristics at baseline, treatment, and outcome measures including treatment response (HES worsening/flare, hematologic relapse, remission), HES-related symptoms, blood/tissue eosinophil counts, OCS and immunosuppressant receipt/reduction in dose, physician-reported organ-related signs and symptoms, treatment switching, health care resource utilization, patient-reported outcomes, and safety.

Study design

Retrospective or prospective observational studies (including database- and registry-based studies), case reports, and case series were included. Animal studies, in vitro studies, genetic studies, randomized controlled trials, systematic reviews, and meta-analyses were excluded.

Sources

On November 1, 2023, we performed searches of the MEDLINE and Embase databases. Searches also included conference abstracts published after January 1, 2021. Search strings are available in Table E2 in the Online Repository at www.jaci-global.org. Further details on sources, screening and selection, data extraction and synthesis, and bias assessment are included in the Methods section in the Online Repository.

Ethics

This study collected data from published studies, so institutional review board approval was not required.

Results

Overall, 206 studies were identified (36 were duplicates); 170 (82.5%) of 206 abstracts were screened, and 87 (51.2%) of 170 were eligible for full-text review. In total, 60 (69.0%) of 87 abstracts/articles were selected for data extraction and synthesis; 5 (8.3%) of 60 were cohort studies, and 55 (91.7%) of 60 were case reports (Fig 1; see Table E3 in the Online Repository at www.jaci-global.org).

Fig 1.

Flow diagram of study-selection process. RCT, Randomized controlled trial.

Cohort studies

Cohort studies characteristics

Of the 5 cohort studies (192 study participants),30, 31, 32, 33, 34 3 were peer-reviewed articles with retrospective designs (169 study participants),30, 31, 32 and 2 were conference abstracts with prospective designs (23 study participants).^33,34 In 3 of the studies, HES was defined as blood eosinophil (bEOS) counts of ≥1500 cells/μL30, 31, 32 and clinical manifestations^30,32 or end-organ damage associated with eosinophilia,³¹ whereas in Aalbers et al³³ (15 study participants), patients were required to have bEOS counts >600 cells/μL and ≥1 affected organ. In Papaioannou et al³⁴ (8 study participants), diagnostic criteria for defining HES were not included.

Patient information at baseline

Of 192 patients, the median (range) age was 44.0 to 61.8 (10.0-86.0) years; 33.3% to 62.9% were female, and 45.5% to 100% had I-HES (Table I).30, 31, 32, 33, 34 Two studies reported on time since diagnosis (n = 46);^30,31 in Kuang et al,³⁰ this was a median (range) of 4.6 (0.4-21.0) years and in Caminati et al,³¹ this was a mean (standard deviation [SD]) of 7.3 (6.4) years. The bEOS counts at baseline were reported in 4 studies (n = 177) and varied from 1000 to 7510 cells/μL.30, 31, 32^,34 All 5 studies reported baseline organ-related signs and symptoms; the most common were pulmonary (50.0-74.0%) and dermatologic (13.0-77.0%; see Fig E1 in the Online Repository at www.jaci-global.org).30, 31, 32, 33, 34

Table I.

Patient demographics and characteristics at baseline in cohort studies

Study measure	Cohort studies (aggregated data)
	Kuang (2018)30	Aalbers (2022)33	Chen (2022)32	Caminati (2023)31	Papaioannou (2023)34
Treatment group	•Mepolizumab monotherapy •Mepolizumab + other treatments •Control (never received mepolizumab)	•Benralizumab 30 mg SC every 4 weeks	•Mepolizumab 100 mg SC •Mepolizumab 300 mg SC •Mepolizumab ≥700 mg IV •Benralizumab∗ •Reslizumab∗	•Mepolizumab 100 mg SC every 4 weeks	•Mepolizumab 300 mg SC every 4 weeks •Benralizumab 30 mg SC every 8 weeks
No. of patients	35†	15	123‡	11	8
Age (years)	Median (range) 44 (12-72)	Median (range) NR (19-69)	Median (range) 45 (10-86)	Median (IQR) 61.8 (55-72)	Mean (SD) 52.1 (20.1)
Female sex	62.9%	33.3%	53.3%	36.3%	45.5%
Time since HES diagnosis (years)	Median (range) 4.6 (0.4-21)	—	—	Mean (SD) 7.3 (6.4)	—
HES subtype
Idiopathic	57.0%	53.3%	45.5%	100%	100%
Lymphocytic	17.0%	40.0%	13.8%	0	0
Myeloid	9.0%	0	0.8%	0	0
Other/overlap	17.0%	7.0%	39.8%	0	0
Index anti–IL-5/anti–IL-5Rα
Mepolizumab	100%	0	83.7%	100%	75.0%
Benralizumab	0	100%	12.2%	0	25.0%
Reslizumab	0	0	4.1%	0	0
bEOS count (cells/μL)	Mean (SD) 3268 (3271)	—	Mean (range) 7510 (1510-89,000)	Median (IQR) 1000 (870-2100)	Mean (SD) 7179 (5599)

IQR, Interquartile range; NR, not reported.

^∗Dose not reported.

^†Patients who received ≥1 dose of mepolizumab and had evaluable data for clinical response.

^‡Total of 117 patients received 123 courses across mepolizumab (n = 103; n = 6 switched to benralizumab), benralizumab (n = 9; n = 0 switched), and reslizumab (n = 5; n = 0 switched).

Study treatment

Four studies assessed mepolizumab in a total of 155 patients,30, 31, 32^,34 3 studies included 32 patients treated with benralizumab,32, 33, 34 and 1 study included 5 patients receiving reslizumab.³² In Kuang et al,³⁰ Caminati et al,³¹ Chen et al,³² and Papaioannou et al,³⁴ most to all patients (75.0-100%) received mepolizumab, which was the only biologic approved at the time of study (Table I). Patients in Kuang et al received mepolizumab 750 mg intravenously (IV; n = 35), patients in Papaioannou et al received mepolizumab 300 mg subcutaneously (SC; n = 6), patients in Caminati et al received mepolizumab 100 mg SC (n = 11),^30,31,34 and patients in Chen et al received mepolizumab 100 mg SC (n = 36), 300 mg SC (n = 37), or ≥700 mg IV (n = 28; not a licensed dose); dose was not reported for patients receiving benralizumab (n = 15) or reslizumab (n = 5).³² Aalbers et al³³ only included patients (n = 15) receiving benralizumab 30 mg SC every 4 weeks (index; 100%), while in Papaioannou et al benralizumab 30 mg SC was administered every 8 weeks (n = 2). Chen et al was the only cohort study to include patients receiving reslizumab (4.1%). Chen et al reported data as number of treatment courses rather than number of patients; any discrepancies in numbers are the result of patients switching between treatment groups and/or doses. Indeed, treatment groups were not mutually exclusive, and patients who switched were counted in each group for which they received a treatment course. Details on prior treatments are included in the Results section and in Fig E2 of the Online Repository at www.jaci-global.org.

Disease worsening/flares

Kuang et al³⁰ compared HES flares between patients with ≥5 years of follow-up who were treated with mepolizumab 750 mg IV for >6 doses (n = 23) and those who did not receive mepolizumab (n = 55). In this study, a HES flare was defined as worsening on >4 weeks of stable treatment and requiring a change in HES medication. Overall, 48% (n = 11) of patients with HES receiving mepolizumab 750 mg IV experienced flares compared with 80% (n = 44) of patients receiving conventional treatment. Patients whose disease was managed by mepolizumab monotherapy experienced significantly fewer mean disease flares (0.193 flares per year) than the control (7.034 flares per year) and patients with disease managed on mepolizumab along with other treatment (0.96 flares per year) (both P < .05). The control cohort included patients receiving conventional treatments other than mepolizumab. In Caminati et al,³¹ none of the 7 patients receiving mepolizumab 100 mg SC experienced a disease flare (defined as worsening of HES-related clinical symptoms or bEOS count increase requiring escalation in therapy that was not related to a decrease in ongoing HES therapy). Similarly, no patients had an asthma flare after treatment with mepolizumab or benralizumab in Papaioannou et al.³⁴

Treatment response

Four studies reported on hematologic or clinical response after anti–IL-5/anti–IL-5Rα treatment.30, 31, 32, 33 In Chen et al,³² of the patients whose disease was not already in hematologic remission at the time of anti–IL-5/anti–IL-5Rα initiation, the disease of >90% of those receiving mepolizumab (n = 51/56) or benralizumab (n = 10/11) showed hematologic remission (bEOS count < 1000 cells/μL); all patients receiving reslizumab (n = 5) were experiencing hematologic remission at the start of treatment. Caminati et al³¹ reported that the disease of all 11 patients with I-HES previously in remission while treated with mepolizumab 100 mg SC successfully maintained remission (control of symptoms and bEOS counts < 500 cells/μL). In Aalbers et al,³³ 86.7% of patients experienced hematologic remission (bEOS counts < 500 cells/μL) after treatment with benralizumab 30 mg SC every 4 weeks (n = 13/15). Among the 35 patients with an evaluable response in Kuang et al,³⁰ the disease of 57.1% showed complete response (improved/resolved symptoms and normal bEOS counts while receiving ≤10 mg prednisone), 20.0% showed partial response (improved symptoms and normal bEOS counts requiring >10 mg prednisone and/or other HES treatment), and 22.9% showed no response. The HES overlap subtype (eosinophilic disease restricted to a single organ system but accompanied by peripheral eosinophilia [bEOS counts >1500 cells/μL]) was the most responsive (100% with complete response, n = 6/6), followed by I-HES (12/20, 60.0%) and lymphocytic HES (2/6, 33.3%).³⁰

Changes in bEOS counts

Over 12 to 48 weeks, reductions in bEOS counts after anti–IL-5/anti–IL-5Rα treatment were observed in 3 studies (n = 54);^30,31,34 bEOS counts were reduced by 53% (from 3268 to 1522 cells/μL) in Kuang et al,³⁰ 100% (from 1000 to <100 cells/μL, P < .001) in Caminati et al,³¹ and 99% (from 7179 to 72 cells/μL) in Papaioannou et al³⁴ (Fig 2).

Fig 2.

Changes in bEOS counts after anti–IL-5/anti–IL-5Rα treatment in cohort studies. Data are expressed as ∗means and SDs and ∗∗medians and interquartile ranges. †73% mepolizumab and 27% benralizumab. Q8W, Every 8 weeks.

Changes in OCS dose

Four studies reported changes in OCS dose after anti–IL-5/anti–IL-5Rα treatment (Fig 3).30, 31, 32^,34 Significant OCS-sparing effects were observed at 12 weeks (15.0 to 4.9 mg/d) for mepolizumab 750 mg IV in Kuang et al,³⁰ and at 12 (10 to 5.0 mg/d, P < .001), 24 (10 to 0 mg/d, P < .001), and 48 weeks (10 to ∼0 mg/d, P = .006) for mepolizumab 100 mg SC in Caminati et al.³¹ In Papaioannou et al,³⁴ a 75% reduction in OCS dose was observed for all 8 patients (mepolizumab/benralizumab). In Chen et al,³² median reductions of 80%, 100%, and 25% were observed for benralizumab (n = 8), mepolizumab (n = 81), and reslizumab (n = 4), respectively. No patients discontinued OCS while receiving reslizumab. For mepolizumab, a 100% median OCS reduction was observed for patients with I-HES (20 to 0 mg/d, n = 34), and a 49% median OCS reduction was observed for patients with lymphocytic HES (17.5 to 6.5 mg/d, n = 12). The percentage reduction in OCS for patients who initiated therapy with mepolizumab ≥ 700 mg IV was similar to that observed with mepolizumab 100 mg SC. A similar pattern was observed for benralizumab: 100% reduction for I-HES (10.5 to 0 mg/d; n = 2) and 60% reduction for lymphocytic HES (20.0 to 8.0 mg/d; n = 1).³²

Fig 3.

OCS receipt before and after anti–IL-5/anti–IL-5Rα treatment in cohort studies. ∗Subset of patients receiving OCS at anti–IL-5/anti–IL-5Rα treatment start. NR, Not reported.

Changes in immunosuppressant dose

Kuang et al³⁰ and Chen et al³² reported on the ability to discontinue, taper, or stop immunosuppressants after anti–IL-5/anti–IL-5Rα treatment (see Fig E3 in the Online Repository at www.jaci-global.org). In Kuang et al, 57.0% of patients discontinued immunosuppressants after treatment with mepolizumab 750 mg IV. In Chen et al, almost all patients receiving immunosuppressants were able to taper or stop after receiving treatment with benralizumab (100%; immunosuppressants were azathioprine and cyclophosphamide) or mepolizumab (83.0%; immunosuppressants were azathioprine, hydroxyurea, IFN-α, methotrexate, and mycophenolate). However, only a quarter could do so after reslizumab (immunosuppressants were hydroxyurea and methotrexate).

Physician-reported organ-related signs and symptoms

Chen et al³² and Aalbers et al³³ reported improvement in physician-reported organ-related signs and symptoms after treatment. In Chen et al, the percentage of patients experiencing physician-reported improvement in overall organ-related signs and symptoms was 67.0% for benralizumab, 60.0% for mepolizumab, and 55.0% for reslizumab (see Fig E4 in the Online Repository at www.jaci-global.org). The authors noted that there appeared to be little correlation between improvement of symptoms with mepolizumab and the initial dose used. Aalbers et al reported that 87% of patients receiving benralizumab (n = 13/15) experienced clinically relevant improvement in organ function after a median of 22 months.

Anti–IL-5/anti–IL-5Rα treatment switching

Nine patients from Chen et al³² and Aalbers et al³³ switched from mepolizumab to benralizumab; reasons were failure of disease to respond (8/9, 88.9%) or an adverse event (AE) where the patient experienced an allergic reaction (1/9, 11.1%). Additionally, one patient in Aalbers et al receiving benralizumab 30 mg SC every 4 weeks stopped treatment because of disease refractory to treatment and switched to mepolizumab. Patient-reported outcomes, along with other long-term clinical outcomes, are reported in the Results section in the Online Repository at www.jaci-global.org.

Safety

Caminati et al³¹ (n = 11) reported no AEs in patients receiving mepolizumab 100 mg every 4 weeks. Similarly, Papaioannou et al³⁴ (n = 8) reported no AEs in a mixed cohort of patients receiving benralizumab 30 mg every 8 weeks (n = 2) or mepolizumab 300 mg every 4 weeks (n = 6). Chen et al³² reported that treatment with mepolizumab (n = 103), benralizumab (n = 15), and reslizumab (n = 5) was well tolerated, but 4 mepolizumab-treated patients had non–life-threatening reactions leading to discontinuation. In Aalbers et al,³³ 2 patients experienced mild headache, and 1 developed a mild malar rash after the first dose of benralizumab 30 mg every 4 weeks (n = 15).

Case reports

Case report characteristics

Of the 55 selected case reports (n = 70), 14 (25.5%) were conference abstracts35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48 and 41 (74.5%) were peer-reviewed articles^1,49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 (Fig 1; Table E3).

Patient demographics/characteristics at baseline

Overall, mean (SD) age was 38.5 (23.7; n = 68) years; 40.0% were female, and 67.2% had I-HES (Table II). Median (interquartile range) time since HES diagnosis was 24.0 (66.0) months (n = 25). All 70 patients reported organ-related signs and symptoms at baseline (see Fig E5 in the Online Repository at www.jaci-global.org).

Table II.

Demographics and characteristics of 70 patients at baseline in case reports

Characteristic	Value
Age at HES diagnosis (years)
No. of subjects	68
Mean [SD]	38.5 [23.7]
Median [IQR]	39.0 [43.5]
Age category at HES diagnosis
0 to <12 years	14 (20.0)
12 to 18 years	7 (10.0)
18 to 34 years	11 (15.7)
35 to 44 years	6 (8.6)
45 to 54 years	8 (11.4)
55 to 64 years	11 (15.7)
≥65 years	11 (15.7)
Missing	2 (2.9)
Sex
Female	28 (40.0)
Male	40 (57.1)
Missing	2 (2.9)
Race
Black	3 (4.3)
White	12 (17.1)
Asian	1 (1.4)
Other	0
Missing	54 (77.1)
Newly diagnosed at time of anti–IL-5/anti–IL-5Rα initiation
Yes	20 (28.6)
No	16 (22.9)
Missing	34 (48.6)
Time (months) from HES diagnosis until anti–IL-5/anti–IL-5Rα initiation
No. of subjects	25
Mean [SD]	45.0 [45.5]
Median [IQR]	24.0 [66.0]
Continent/country of residence
North America	35 (50.0)
Europe	27 (38.5)
Israel	2 (2.9)
Africa	1 (1.4)
Asia	3 (4.3)
Australia	1 (1.4)
Missing	1 (1.4)
Distribution of HES subtypes
Idiopathic	39 (67.2)
Lymphocytic	16 (27.6)
Myeloproliferative	1 (1.7)
Immune checkpoint inhibitor–induced HES	2 (3.5)
Patients with specific comorbidities
Any comorbidities	39 (55.7)
Asthma	22 (31.4)
Chronic rhinosinusitis	9 (12.9)
Diabetes	4 (5.7)
Nasal polyps	4 (5.7)
Hypertension	4 (5.7)
Cerebrovascular disease	4 (5.7)
Anxiety or depression	3 (4.3)
Eosinophilic granulomatosis with polyangiitis	2 (2.9)
Obesity	2 (2.9)
Osteoporosis	1 (1.4)
Patients with organ involvement
Any manifestations or organ involvement	70 (100)
Dermatologic	39 (55.7)
Pulmonary	34 (48.6)
Constitutional (general)	27 (38.6)
Cardiovascular	18 (25.7)
Gastrointestinal/bladder	18 (25.7)
Musculoskeletal/connective	14 (20.0)
Vascular	13 (18.6)
Neurologic	10 (14.3)
Otolaryngologic	9 (12.9)
Hepatobiliary	5 (7.1)
Renal	3 (4.3)
Ocular	2 (2.9)

Data are presented as nos. (%) unless otherwise indicated. IQR, Interquartile range.

Treatment

Forty-four case reports provided data on treatment outcomes after mepolizumab,^1,37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54^,57, 58, 59, 60^,63, 64, 65^,67,69,72,74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88 8 after benralizumab,^{35,36,61,62,66,68,70,71} and 3 after reslizumab.^55,56,73 Of the 70 patients who received anti–IL-5/anti–IL-5Rα treatment, 54 (77.1%) started on mepolizumab therapy,^1,37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54 13 (18.6%) on benralizumab,^{1,35,36,61,62,66,68,70,71} and 3 (4.3%) on reslizumab.^55,56,73 Dosing of mepolizumab varied between case reports, with 35.2% (n = 19/54) of patients having received the approved dose of 300 mg SC. Information on dose was missing for 24.1% (n = 13/54) of patients receiving mepolizumab and 23.1% (n = 3/13) of patients receiving benralizumab (see Table E4 in the Online Repository at www.jaci-global.org). Reasons for starting anti–IL-5/anti–IL-5Rα treatment included inadequate response to previous treatments or disease progression despite treatment (n = 24, 64.9%); HES-related symptom management (n = 24, 64.9%); OCS-sparing strategy, adverse effects, and safety (n = 5, 13.5%); insurance and access issues (n = 1, 2.7%); and other reasons including switching clinics and starting anti–IL-5/anti–IL-5Rα as an outpatient (n = 2, 5.4%). Details on prior treatments are included in the Results section in the Online Repository.

Disease worsening/flare

HES worsening/flare was documented in 13 patients (12 patients [22.2%] receiving mepolizumab and 1 patient [7.7%] receiving benralizumab), defined as follows: initiation of short-course OCS; receipt of IV methylprednisolone; initiation of OCS/increase in OCS dose of ≥10 mg/d; decrease in dosing for existing SC/IV immunosuppressants; initiation of a new immunosuppressant (excluding change after cyclophosphamide receipt); or a HES-related hospitalization or emergency department visit for eosinophilia and HES symptoms.^{1,37,41,45,49,54,60,}64, 65, 66, 67^,80,82

Treatment response

Nearly all patients (98.6%) had improvement of overall HES symptoms. Only 5 patients (7.1%) experienced hematologic relapse (bEOS counts ≥ 1000 cells/μL) after mepolizumab treatment,^{1,59,60,67,88} 3 of whom discontinued treatment.^59,60,88

Changes in bEOS counts

Of 37 patients who reported changes in bEOS counts, significant reductions were observed. In many cases, a 95% reduction in bEOS counts was reported: mean (SD) 7444.9 (15,128.3) cells/μL at baseline to 408.9 (1263.5) cells/μL after anti–IL-5/anti–IL-5Rα treatment initiation (P = .0066) over a mean follow-up of 39.0 weeks. In patients evaluated for reductions in bEOS count, 10 received benralizumab, 26 mepolizumab, and 1 reslizumab. Mean (SD) bEOS counts were reduced from 7410.0 (11,866) cells/μL at baseline to 58.0 (183.4) cells/μL at follow-up (P = .082) for benralizumab, and from 7711.1 (16,646) cells/μL at baseline to 559.6 (1485.6) cells/μL at follow-up (P = .034) for mepolizumab. The bEOS reductions were not calculated for reslizumab.^{39,43,46,50,52,53,}56, 57, 58^,60, 61, 62, 63^,65, 66, 67^,69, 70, 71, 72^,76,78,81, 82, 83^,85, 86, 87, 88

Changes in OCS and IV steroid dose

Mean (SD) daily OCS dose decreased significantly by 89%, from 35.8 (24.4) to 3.8 (6.9) mg/d (P < .0001), over a mean follow-up of 27.0 weeks.^{1,40,51,52,56,58,59,61,63,66,68,69,71,72,76,79,}83, 84, 85, 86 Among those with available data (n = 35/70), 82.9% of patients had a 50% reduction in OCS dose.^1,35,43,51, 52, 53^,56,58,61, 62, 63^,66,68,69,71, 72, 73^,76,79,82,84, 85, 86

In a subgroup analysis, data on daily OCS dose during anti–IL-5/anti–IL-5Rα treatment were available for 4 of 13 patients receiving benralizumab and 18 of 54 patients receiving mepolizumab. The mean daily pretreatment OCS dose for patients receiving benralizumab was 40.0 mg/d (SD 16.3), which decreased to 0 mg/d (SD 0.0) after treatment (P = .0163). Patients receiving mepolizumab had a mean (SD) pretreatment daily OCS dose of 32.5 (24.4) mg/d, which decreased to 4.3 (7.5) mg/d after treatment (P = .0003). Data were not presented for patients receiving reslizumab because of limited data. A greater proportion of patients receiving benralizumab (10/13, 76.9%) experienced an OCS-sparing effect, with a ≥50% reduction in their OCS dose compared with those receiving mepolizumab (17/54, 31.5%). Information on IV steroid dose during anti–IL-5/anti–IL-5Rα treatment was documented for 2 patients; it was reduced from 50.0 (SD 14.1) mg to 2.5 (SD 3.5) mg/d.^70,71

Physician-reported organ-related signs and symptoms

Thirty-eight patients (55.1%) with physician-reported organ-related signs and symptoms at baseline had improvement after anti–IL-5/anti–IL-5Rα treatment (see Fig E6 in the Online Repository at www.jaci-global.org).^{39,41,42,44,46,47,}49, 50, 51, 52, 53^,55, 56, 57, 58, 59^,63,66,68, 69, 70^,74, 75, 76, 77, 78, 79, 80^,82,85, 86, 87, 88 After initiating treatment, new organ-related signs and symptoms were only observed in those treated with mepolizumab (7/54, 13.0%).^{1,45,58,59,79,88}

Anti–IL-5/anti–IL-5Rα treatment switching

Six patients (8.6%) switched from mepolizumab therapy to benralizumab.^61,62,70 Reasons for switching included partial control of asthma but HES relapse (n = 2, 33.3%), no control of asthma (n = 2, 33.3%), HES relapse and increase in OCS requirement (n = 1, 16.7%), and HES flare (n = 1, 16.7%). One patient (1.4%) switched from benralizumab to mepolizumab because of eosinophilia.¹ One patient (1.4%) switched from mepolizumab to reslizumab after experiencing a hematologic relapse.⁷³ One patient (1.4%) switched from reslizumab to mepolizumab for a reason not specified.⁵⁹

Of the 6 patients (8.6%) who switched from mepolizumab to benralizumab, all experienced normalization of bEOS counts, and 83% (n = 5/6) experienced OCS dose reductions. One patient (16.7%) experienced clinical remission, which was maintained for 2 years.^61,62,70

In the patient (1.4%) who switched from benralizumab to mepolizumab, an improvement in fatigue, swelling, and severe pruritis was observed. In this patient, OCS dose was tapered from 60 mg to 9 mg.¹

In the patient (1.4%) who switched from mepolizumab to reslizumab, bEOS counts were maintained below 500 cells/μL and OCS dose was reduced.⁷³

In the patient (1.4%) who switched from reslizumab to mepolizumab, no clinical response was observed, and mepolizumab was subsequently stopped because of its lack of efficacy.⁵⁹

Anti–IL-5/anti–IL-5Rα discontinuation (switching not included)

Of the 54 patients who started on mepolizumab therapy, 13 (24.1%) discontinued it.^{37,52,59,60,65,}79, 80, 81^,84,88 Of these 13 patients, 11 had data on the duration of mepolizumab therapy. Nearly all (10/11, 90.9%) discontinued within 12 months. Reasons for discontinuation included no insurance coverage (n = 1, 7.7%),⁵² resolution of hypereosinophilia (n = 1, 7.7%),⁸⁴ lack of efficacy or minimal response to mepolizumab (n = 8, 61.5%),^{37,59,60,65,80,88} serious AEs involving lymphoma (n = 1, 7.7%),⁵⁹ and unknown (n = 2, 15.4%).^79,81 No patients discontinued benralizumab or reslizumab. Patient-reported outcomes, other long-term clinical outcomes, and safety are reported in the Results section in the Online Repository at www.jaci-global.org.

Discussion

At present, mepolizumab 300 mg SC is the only approved anti–IL-5 treatment for HES, with no anti–IL-5Rα treatments currently approved for patients with HES;^20,21 as such, data on the use of these treatments in the real world are limited. This SLR extracted data from case studies and a small number of observational studies to consolidate information and provide a greater understanding on the real-world use of anti–IL-5/anti–IL-5Rα treatments and their clinical outcomes in patients with HES. In this SLR, anti–IL-5/anti–IL-5Rα treatments were associated with achieving important clinical objectives such as reductions in OCS requirement, reduction in bEOS counts, and improvement in physician-reported organ-related signs and symptoms in both cohort studies and case reports. Furthermore, nearly all patients in cohort studies experienced hematologic remission after treatment with benralizumab or mepolizumab, and hematologic relapses were uncommon (7.1%) in case studies.^1,30, 31, 32, 33^,59,60,67,88

Treatment with OCS is associated with serious risks,^12,14, 15, 16, 17 and therefore approaches that maintain efficacy while minimizing OCS dose are considered beneficial. All cohort studies reported reductions in OCS dose with anti–IL-5/anti–IL-5Rα treatments.30, 31, 32, 33, 34 In one study, reductions in OCS dose were greater in patients with I-HES versus lymphocytic HES,³² which is consistent with previous research.⁸⁹ Moreover, a reduction in mean daily OCS dose of almost 90% was reported across case reports.

Reduction in bEOS count is an important treatment goal for the prevention of eosinophil-related organ damage and clinical sequelae.¹¹ Analyses showed that many patients had an almost complete depletion of eosinophils with anti–IL-5/anti–IL-5Rα treatment. In case reports, mepolizumab reduced mean bEOS counts by ∼93% (mean [SD] 7711.1 [16,646] cells/μL at baseline to 559.6 [1485.6] cells/μL at follow-up; P = .034). The greatest reductions in bEOS counts reported in case reports (∼99%) was observed in patients who received benralizumab; however, the reduction did not achieve statistical significance because of the small number of patients receiving this treatment (n = 10) (mean [SD] 7410.0 [11,866] cells/μL at baseline to 58.0 [183.4] cells/μL at follow-up; P = .082).

After anti–IL-5/anti–IL-5Rα treatment, hematologic remission was maintained or obtained in 86.7% to 100% of patients from cohort studies, which is aligned with the response rate observed in patients from case reports. However, remission and response definitions varied among the studies, making it challenging to compare the differences in remission/response rates.

Results from case reports showed that more than half (55%) of patients experienced improvement in physician-reported organ-related signs and symptoms after anti–IL-5/anti–IL-5Rα treatment, whereas in cohort studies, overall improvements were observed in 67% to 87% of patients treated with benralizumab, 60% of patients treated with mepolizumab, and 55% of patients treated with reslizumab. Differences in improvement may be explained by the variance in median follow-up.

The SLR is subject to various limitations. Although bias was minimized by combining all patient-level data from case reports into a database so they could be analyzed and compared with the larger cohort studies, a formal bias assessment was not conducted. Therefore, biases are likely present and may influence generalizability of the findings. For example, searches were only conducted for English-language articles indexed in MEDLINE, Embase, and select international conferences, and other relevant publications were not captured. Until recently, there has been a lack of a specific International Classification of Diseases, Tenth Revision, Clinical Modification, code for HES, which is not yet in use in many European countries.^3,90, 91, 92 This may have also limited the inclusion of studies from some regions, representing another potential bias. Most publications included in this SLR were case series or case reports, so findings may not be generalizable to the wider HES population because of their lack of a controlled study design and risk of selection bias.⁹³ The use of case reports may result in a bias toward reporting of rare clinical presentations or patients with high disease severity. Indeed, nearly one third of patients from case reports were children, even though most patients with HES are usually diagnosed in adulthood.¹¹ This SLR was heavily weighted toward patients receiving mepolizumab, which was the only approved anti–IL-5 treatment for patients with HES at the time of analysis.^20,21 Both the case reports and the cohort studies highlighted that patients were receiving a variety of mepolizumab doses. Of patients who received mepolizumab, approximately 35% (n = 19/54) of patients in the case reports and 28% (n = 43/155) (Papaioannou et al³⁴ [n = 6] and Chen et al³² [n = 37]) of patients in the cohort studies used the mepolizumab dose and route of administration approved for HES (300 mg SC).⁹⁴

Benralizumab was not approved for HES at the time of SLR conduct; thus, most studies only reported on the approved asthma dosing for benralizumab (30 mg SC every 8 weeks),^95,96 with the exception of the Aalbers et al³³ study, which used an off-label every-4-weeks dosing frequency. Nevertheless, these studies provide an early insight into the real-world benefits of benralizumab in patients with HES. The differences in doses across treatments may introduce some variability in responses, making indirect comparisons and interpretation of the data challenging. Additional biases may also arise from the SLR because one cohort study enrolled over 100 patients and contributed two thirds of patients to the analyses,³² whereas other cohort studies enrolled only small samples.^30,31 Nevertheless, the smaller cohort studies contributed useful information, with two increasing the amount of data available for benralizumab.^33,34 Furthermore, because of differences in reporting between studies, the data included in the selected publications were not aligned. For example, in case reports, mean follow-up could not be calculated for the patients who obtained a 50% reduction in OCS dose. Finally, the bEOS diagnostic criteria of >600 cells/μL in one study³³ differed from other studies30, 31, 32 and was not in line with the generally accepted cut-off for HES (≥1500 cells/μL),^10,11 further limiting the generalizability of those data. However, key outcomes such as OCS reduction, changes in bEOS counts, and rates of HES worsening/flares were generally similar between studies.

Conclusion

To our knowledge, our study is one of the few SLRs to provide a comprehensive overview of patients with HES receiving anti–IL-5/anti–IL-5Rα treatment in real-world practice. Despite potential biases and limitations, the studies included in this SLR offer an opportunity to assess the real-world use of anti–IL-5/anti–IL-5Rα treatment in a broader population of patients with HES than would be eligible for, or participate in, clinical trials, and to provide early insights into the effects of newer therapies such as benralizumab. Although small sample sizes were not conducive to direct comparisons between therapies, our SLR demonstrates that monoclonal antibody therapies targeting eosinophils may improve outcomes for a broad range of patients with HES in the real world. Anti–IL-5/anti–IL-5Rα treatment was associated with OCS-sparing effects without exacerbating HES symptoms or inducing hematologic relapse. Further studies with larger populations are needed to better explore differences in the clinical outcomes that are based on the type of anti–IL-5/anti–IL-5Rα treatment utilized to treat HES.

Clinical implication.

This comprehensive SLR of observational cohort studies and case reports advances understanding of the real-world use and outcomes of anti–IL-5/anti–IL-5Rα therapies in patients with HES.

Disclosure statement

Funded by AstraZeneca. Evidera received research funding/consulting fees from AstraZeneca for the design and implementation of analyses.

Off-label disclosure: This publication contains information from case studies, case series, and cohort studies on the real-world use and impact of benralizumab, mepolizumab, and reslizumab in patients with hypereosinophilic syndrome (HES). At the time of analysis, benralizumab and reslizumab were not approved for the treatment of HES, and intravenous doses of mepolizumab were considered off-label.

Disclosure of potential conflict of interest: S. Y. Chen, C. Edmonds, and P. Jain report employment at and holder of stock/shares in AstraZeneca. N. Ding and M. Stokes report employment at Evidera, which received consulting fees for this work from AstraZeneca. J. Rowell reports employment at and holder of stock/shares at AstraZeneca; and current holder of stock/shares at Roche.