Efficacy and safety of gluten peptide-based antigen-specific immunotherapy (Nexvax2) in adults with coeliac disease after bolus exposure to gluten (RESET CeD): an interim analysis of a terminated randomised, double-blind, placebo-controlled phase 2 study

Background: A gluten-free diet is insufficient to treat coeliac disease because intestinal injury persists and acute reactions with cytokine release follow gluten exposure. Nexvax2 is a specific immunotherapy using immunodominant peptides recognised by gluten-specific CD4⁺ T cells that might modify gluten-induced disease in coeliac disease. We aimed to assess the effects of Nexvax2 on gluten-induced symptoms and immune activation in patients with coeliac disease.

Methods: This was a randomised, double-blind, placebo-controlled phase 2 trial done at 41 sites (29 community, one secondary, and 11 tertiary centres) in the USA, Australia, and New Zealand. Patients with coeliac disease aged 18-70 years who had excluded gluten for at least 1 year, were HLA-DQ2.5 positive, and had a worsening of symptoms after an unmasked 10 g vital gluten challenge were eligible for inclusion. Patients were stratified by HLA-DQ2.5 status (HLA-DQ2.5 non-homozygous vs homozygous). Patients who were non-homozygous were centrally (ICON; Dublin, Ireland) randomly assigned (1:1) to receive subcutaneous Nexvax2 (non-homozygous Nexvax2 group) or saline (0·9% sodium chloride; non-homozygous placebo group) twice a week escalating from 1 μg to 750 μg during the first 5 weeks followed by 11 weeks of maintenance therapy at 900 μg per dose. The exploratory homozygous group was centrally randomly assigned (2:1) to receive Nexvax2 (homozygous Nexvax2 group) or placebo (homozygous placebo group); patients who were homozygous received the same dosage as those who were non-homozygous. The primary endpoint was change in coeliac disease patient reported outcomes (total gastrointestinal domain) from pretreatment baseline to the day of masked bolus 10 g vital gluten challenge given in week 14 analysed in the non-homozygous intention-to-treat population. The trial is registered with ClinicalTrials.gov, NCT03644069.

Findings: Between Sept 21, 2018, and April 24, 2019, 383 volunteers were screened for inclusion, of whom 179 (47%; 133 [74%] women, 46 [26%] men; median age 41 years [IQR 33-55]) were randomly assigned. One (1%) of 179 patients was excluded from analysis due to misassignment of genotype. The non-homozygous Nexvax2 group included 76 patients, the non-homozygous placebo group included 78 patients, the homozygous Nexvax2 group included 16 patients, and the homozygous placebo group included eight patients. The study was discontinued after planned interim analysis of 66 patients who were non-homozygous. We report an unmasked post-hoc analysis of all available data for the primary endpoint and secondary symptom-based endpoints combining data from 67 (66 were assessed in the planned interim analysis for the primary endpoint). Mean change from baseline to day of first masked gluten challenge in total gastrointestinal score for the non-homozygous Nexvax2 group was 2·86 (SD 2·28) compared with 2·63 (2·07) for the non-homozygous placebo group (p=0·43). Adverse events were similar between all patients who received Nexvax2 and those who received placebo. Serious adverse events were reported in five (3%) of 178 patients (two [2%] of 92 who received Nexvax2 and three [4%] of 82 who received placebo). One patient in the non-homozygous Nexvax2 group had a serious adverse event that occurred during gluten challenge (left-sided mid-back muscle strain with imaging suggestive of partial left kidney infarction). Serious adverse events were reported for three (4%) of 78 patients in the non-homozygous placebo group (one each with exacerbation of asthma and appendicitis, and one who had forehead abscess, conjunctivitis, and folliculitis) and one (1%) patient in the non-homozygous Nexvax2 group developed a pulmonary embolism. The most frequent adverse events in all 92 patients who received Nexvax2 compared with all 86 patients who received placebo were nausea (44 [48%] of 92 patients who received Nexvax2 vs 29 (34%) of 86 patients who received placebo), diarrhoea (32 [35%] vs 25 [29%]), abdominal pain (31 [34%] vs 27 [31%]), headache 32 [35%] vs 20 [23%]), and fatigue (24 [26%] vs 31 [36%]).

Interpretation: Nexvax2 did not reduce acute gluten-induced symptoms. Masked bolus vital gluten challenge provides an alternative to extended gluten challenge in efficacy studies for coeliac disease.

Funding: ImmusanT.