Dupilumab versus placebo in adults and adolescents with eosinophilic gastritis (DEGAS): a double-blind, placebo-controlled, phase 2, multicentre, randomised controlled trial.

backgroundEosinophilic gastritis currently has no approved treatments and is postulated to be driven by type 2 inflammation. Dupilumab blocks type 2 cytokines IL-4 and IL-13 and has efficacy in multiple diseases characterised by type 2 inflammation, including eosinophilic oesophagitis. We aimed to assess the efficacy and safety of dupilumab in patients with eosinophilic gastritis.

methodsDEGAS was a proof-of-concept, phase 2, multicentre, randomised controlled trial consisting of a 12-week, double-blind, placebo-controlled period, followed by a 24-week open-label extension period. Patients aged 12-70 years from 11 hospitals in the USA with histologically active eosinophilic gastritis (≥30 eosinophils per high-power field [HPF] in at least five HPFs in the gastric antrum and/or body) and moderate-to-severe symptoms occurring at least 2 days per week in the 2 weeks before screening were recruited. Patients with current or recent use of any biologic or current use of systemic steroids at a dose of more than 10 mg/day (prednisone) were excluded. Eligible patients were individually randomised (1:1) to parallel groups and received six injections over 12 weeks: subcutaneous dupilumab (600 mg once followed by 300 mg every 2 weeks) or subcutaneous placebo. Randomisation was performed using a central variable block (block sizes permuted between 2 and 4), with stratification by age (12-17 years or ≥18 years) and use (yes or no) of either systemic corticosteroids, swallowed corticosteroids for eosinophilic gastritis, or non-steroidal systemic immunosuppression therapy. Throughout the duration of the study, patients were expected to maintain their treatments or diets for eosinophilic gastritis. All patients who completed the double-blind period could enter the open-label extension at week 12, during which both groups received dupilumab until week 34. The primary endpoint of relative change from baseline in mean gastric eosinophil count from the five most eosinophil-dense HPFs in the gastric antrum and/or body was analysed at week 12 using linear regression. Secondary endpoints included absolute changes from baseline in Eosinophilic Gastritis Histologic Scoring System (EoS-HSS) total score, mean gastric eosinophil count from the five most eosinophil-dense HPFs, and Eosinophilic Gastritis Endoscopic Reference System (EoG-REFS) total score. All randomly assigned patients who received at least one dose of study drug were included in the safety analysis and efficacy analyses were done in all randomly assigned patients who received at least one dose of study drug and had outcome data available at week 12 (complete case). This study is registered with ClinicalTrials.gov (NCT03678545) and is now complete.

findingsBetween May 14, 2021, and Nov 10, 2023, we randomly assigned 41 patients, of whom 21 (51%) received dupilumab and 20 (49%) received placebo during the double-blind period and were included in the safety analysis. Patients were aged 12-59 years (mean 30·5 years [SD 13·2]; seven [17%] aged <18 years), 37 (90%) were White, one (2%) was Asian, one (2%) was Black or African American, two (5%) were of multiple races, 25 (61%) were female, and 16 (39%) were male. One patient from the placebo group withdrew before week 12; the remaining 21 patients treated with dupilumab and 19 patients treated with placebo had available data and were assessed for the primary endpoint. At week 12, the relative reduction in the primary endpoint was greater with dupilumab (estimated mean change -50% [95% CI -66 to -34]) than with placebo (-4% [-20 to 13]; difference -47 percentage points [-70 to -24]; p<0·0001). Significant differences between groups were noted for the secondary endpoints of absolute change from baseline in EoS-HSS total score (difference -0·10 [95% CI -0·18 to -0·03]; p=0·0055), absolute change from baseline in mean gastric eosinophil count (-38·8 [-75·6 to -17·8]; p=0·0008), and absolute change from baseline in EoG-REFS total score (-3·42 [-6·18 to -0·65]; p=0·016). At week 12, the incidence of treatment-emergent adverse events was similar between dupilumab (17 [81%]) and placebo (17 [85%]). The most common adverse event was blood eosinophilia, with similar incidence in the dupilumab (six [29%]) and placebo (six [30%]) groups. No serious adverse events or treatment-related deaths were reported.

interpretationThe improvement of histological outcomes of eosinophilic gastritis with dupilumab in this proof-of-concept study shows type 2 inflammatory involvement in the disease and the potential value of dupilumab for the treatment of eosinophilic gastritis.

fundingNational Institutes of Health, USA; Regeneron Pharmaceuticals Inc; and Sanofi.