Background/Purpose: Low drug levels of anti-Tumor Necrosis Factor (TNF) therapies are related to poorer response in patients with rheumatoid arthritis (RA). Furthermore, anti-drug antibodies (ADA) against several TNF-inhibitors have been described and are associated with lower drug levels and therefore with lower response rates in RA.

Objectives: To study the relation between clinical response, golimumab levels and ADA in RA patients treated with golimumab.

Methods: Prospective observational cohort study of 38 consecutive RA patients treated with golimumab 50 mg subcutaneously (SC) once every month in the Netherlands (n=36) and Spain (n=2), monitored during 28 weeks of follow-up. Golimumab levels and ADA titres were determined retrospectively using an enzyme linked immunosorbent assay and an Antigen Binding Test (ABT), respectively, designed by Sanquin Research, Amsterdam. Response was defined according to the European League Against Rheumatism (EULAR) response criteria. The last observation was carried forward for patients who dropped out before week 28.  

Results: At baseline, most patients had an active disease (median DAS28 (IQR) 4.5 (3.1-5.5)), 24 patients used concomitant methotrexate (median mg/ week (IQR)) (11.3 (0-25)) and 18 used prednisone (median mg/day (IQR)) (7.5 (5-10)). Fourteen patients had used 1 prior biological (anti-TNF) and 10 patients had used 2 or more prior biologicals (anti-TNF and/or another). At 28 weeks, 10 patients were good, 9 were moderate and 19 were EULAR non-responders. In total, 16 patients dropped-out before week 28; 11 due to inefficacy and 5 due to adverse events.

Median golimumab level at 28 weeks of treatment (µg/mL)(IQR) was 0.8 (0.3-1.9) and varied from undetectable to 3.6µg/mL. Median golimumab levels were lower in non-responder (0.4 (0.2-1.5)) vs. moderate (1.0 (0.2-1.9)) and good responders (1.6 (0.7-2.4)) at 28 weeks of treatment, although this difference was not statistically significant (p=0.08). In 2 patients high ADA titres (between 303 and 1510 AU/mL) were detectable at week 16 which resulted in undetectable golimumab levels. In 1 other patient ADA were detectable (36.8 AU/mL) at week 4 which resulted in a low level of  golimumab (0.35 µg/mL). All 3 patients were EULAR non-responder and golimumab was discontinued before 28 weeks of treatment due to inefficacy according to the treating rheumatologist.

Conclusion: In our study of 38 RA patients, golimumab levels were lower in EULAR non-responders compared to moderate and good responders at 28 weeks of treatment. In 3 patients ADA were detectable which resulted in undetectable golimumab levels in two and low golimumab levels in one patient. All ADA positive patients dropped-out before 28 weeks of treatment due to inefficacy.  

Figure: Median golimumab levels (µg/mL) (last observation carried forward) for EULAR good, moderate and non-responders at 28 weeks of treatment (p=0.08).