Comparison of the ELISA and ECL Assay for Vedolizumab Anti-drug Antibodies: Assessing the Impact on Pharmacokinetics and Safety Outcomes of the Phase 3 GEMINI Trials

Abstract

Vedolizumab immunogenicity has been assessed using an enzyme-linked immunosorbent assay (ELISA) with a ~ 0.5 μg/mL drug interference, which may underestimate on-drug immunogenicity. We aimed to compare immunogenicity results between ELISA and the new drug-tolerant electrochemiluminescence (ECL) assay (and the two versions of neutralizing assays, drug-sensitive versus drug-tolerant). The ECL assay drug tolerance is ~ 100 times higher than that of the ELISA (≥ 50 μg/mL vs. 0.5 μg/mL with a 500 ng/mL positive control), and assay sensitivity is < 5 ng/mL for both assays. Vedolizumab immunogenicity was assessed in 2000 GEMINI 1 and 2 patients originally tested by ELISA and retested by ECL assay. Anti-drug antibody (ADA) impact on infusion-related reactions and pharmacokinetics (PK) was examined using descriptive statistics and population PK analyses. By ECL assay, 6% (86/1427) of patients treated with vedolizumab as induction and maintenance therapy tested ADA-positive. Of these, 20 patients were persistently positive and 56 had neutralizing antibodies. By ELISA, 4% (56/1434) of these patients were ADA-positive, 9 were persistently positive, and 33 had neutralizing antibodies. Among 61 patients with infusion-related reactions, 6 (10%) were ADA-positive (2 persistently positive) by ECL assay. By ELISA, 3 (5%) patients were both ADA-positive and persistently positive. Most results (96%) were similar with both assays. In the updated population PK model, ADA-positive status was estimated to increase vedolizumab linear clearance by a factor of 1.10 (95% credible interval 1.03-1.17), which is consistent with previous reports. The impact of ADA on safety and PK modeling remained generally consistent using either ELISA or ECL assay. ClinicalTrials.gov: NCT00783718 and NCT00783692.

Keywords: ELISA; electrochemiluminescence; immunogenicity; vedolizumab.

Conflict of interest statement

Timothy Wyant and Maria Rosario were employees of Takeda at the time this research was conducted. Lili Yang is an employee of Takeda and holds Takeda stocks or stock options.

Figures

Fig. 1

Modeled effect of covariates on vedolizumab linear clearance (CLL) using the ECL assay data. CLL for a patients with UC and b patients with CD relative to a typical reference patient (UC referent 70 kg, albumin 4 g/dL, fecal calprotectin 700 mg/kg, partial Mayo score 6, 40 years old, naive anti-TNFα therapy, ADASUB-negative, and no adjuvant azathioprine, mercaptopurine, methotrexate, or aminosalicylate therapy; CD referent 70 kg, albumin 4 g/dL, fecal calprotectin 700 mg/kg, CDAI score 300, 40 years old, anti-TNF therapy naive, ADA-negative, and no adjuvant azathioprine, mercaptopurine, methotrexate, or aminosalicylate therapy) is plotted by covariate value. Covariates were fixed to the reference values except when they were the subject of perturbation. Body weight and albumin were evaluated at the observed 5th, 25th, 75th, and 95th percentiles in the data set. The closed circles represent the median and the horizontal lines represent the derived 95% CDI. The vertical dashed line at x = 1 represents the typical reference patient, and the grey-shaded region represents a parameter change of ± 25% from the reference value of 1 (null effect). ADA vedolizumab anti-drug antibody, ADASUB patient-level vedolizumab ADA incidence indicator, CD Crohn’s disease, CDAI Crohn’s Disease Activity Index, CDI credible interval, ECL electrochemiluminescence, TNF tumor necrosis factor alpha, UC ulcerative colitis