Background: Golimumab is an anti-TNF antibody with higher affinity for TNF compared to infliximab and adalimumab and higher stability compared to infliximab1. The PURSUIT trials showed a significant exposure-response relationship with golimumab in ulcerative colitis (UC). Low drug exposure may be related to primary and secondary non-response as described for infliximab and adalimumab. To assess if assay results obtained with different methods can be considered comparable, we conducted a comparison of different immunoassays measuring levels of golimumab and anti- drug antibodies (ADA).
Methods: We included serum samples from 80 patients with UC recruited in an ongoing prospective observational study who started on golimumab treatment. Golimumab was quantified by either an anti-IgG detection antibody (Theradiag) or an antibody directed against the idiotype of golimumab (Sanquin and KU Leuven, Janssen R&D). Bridging drug-sensitive ELISA assays (Theradiag, Janssen R&D, KU Leuven), a bridging drug-tolerant ELISA assay (Janssen R&D), and a radioimmunoassay (Sanquin) were used to quantify ADA.
Results: Median (IQR) serum golimumab levels were 4.5 (2.2–6.6), 3.5 (1.7–4.8), 4.9 (2.6–7.0), and 2.4 (1.2–4.0) μg/mL with Theradiag, Sanquin, KU Leuven and Janssen R&D assays, respectively. Median assay values are different between assays in head to head comparisons (p=0.0001), except for comparison of median levels between Theradiag and KU Leuven (p=0.155) and comparison of Sanquin and Janssen R&D assay values (p=0.055). With the Sanquin and Leuven assays, 84% of samples are in the same quartile of distribution of values; this overlap was 71% for Sanquin and Theradiag, 66% for Theradiag and Leuven assays, 68% for Theradiag and Janssen R&D, 81% for Leuven and Janssen R&D and 86% for Sanquin and Janssen R&D. All but not the Theradiag assays showed specificity for golimumab as these make use of an anti-idiotype antibody to golimumab for detection. We were not able to compare ADA concentrations determined by the different assays. However, ADA were detected in the two same patients by Theradiag, Sanquin and KU Leuven assays and in one of those two with the drug-sensitive Janssen R&D assay. ADA were detected in 27.5% of samples (22/80) with the Janssen R&D drug-tolerant assay.
Conclusions: There is an acceptable correlation between golimumab serum concentrations measured with Theradiag, Sanquin, KU Leuven and Janssen R&D assays. The low number of ADA positive samples precludes statistical concordance between ADA values measured with the different assays.