Objectives To study the correlation between levels of the anti-TNF biologics and clinical efficacy in patients with inflammatory arthritis

Methods Adult patients who fulfilled the criteria for rheumatoid arthritis (RA), spondyloarthritis (SpA) and psoriatic arthritis (PSA) and were commenced on standard doses of the anti-TNF biologics were recruited. Serum samples saved at baseline, month 6 and 12 were assayed for the trough levels of the biologics (± anti-drug antibodies) retrospectively. Patients were followed longitudinally and efficacy analyses were conducted at 3-month intervals without the knowledge of the drug levels. Biologics would be discontinued from 6 months onwards according to protocol-based improvement criteria for each disease. Clinical efficacy of the anti-TNF biologics was compared among patients with different levels of the drug by statistical methods.

Results 112 patients were studied (58 RA, age 51.2±10.9 years, disease duration 72.9±67 months; 41 SpA, age 39.1±9.9 years, disease duration 74.3±81.6 months; 13 PSA, age 53.5±10.7 years, disease duration 44.3±35.4 years). The number of patients treated with infliximab (IFX), adalimumab (ADM), golimumab (GLM) and etanercept was 3, 31, 36 and 42, respectively. At month 12, neutralizing antibodies against IFX, ADM and GLM were present in 2 (67%), 14 (45%) and 1 (3%) of the patients, respectively. In ADM users, the drug level was significantly lower in those with antibodies than those without (1.81±2.63 vs 8.02±4.14 ug/ml; p<0.001). Antibody titer against ADM correlated negatively with the levels of ADM (Rho -0.72; p<0.001). Patients were stratified arbitrarily into 3 groups for each biologic according to the trough levels of the drugs. Low drug concentrations were defined as levels ≤1.30 ug/ml, 0.05 ug/ml and 0.60 ug/ml in ADM, ETN and GLM users, respectively. In patients with RA/PSA (N=71), patients with the lowest anti-TNF drug level group (N=30) had a non-significant trend of less improvement in DAS28, CDAI scores at month 12 when compared to others (N=41). However, significantly more patients withdrew treatment due to inefficacy at month 12 in this group compared to others (67% vs 7.3%, p<0.001). In patients with SpA (N=41), patients with lowest anti-TNF drug levels stratum (N=9) had significantly less improvement in ASDAS compared with others at month 12 (N=32) (-0.57±0.63 vs -1.93±1.28; p=0.003). The proportion of patients who achieved an ASAS20 response was also significantly lower in this group of patients (33% vs 75%; p=0.04). In all the 112 patients studied, the cumulative withdrawal rate of the anti-TNF biologics at month 12 (by Kaplan-Meier's analysis) was significantly higher in those with low drug levels when compared to others (26.1% vs 54.6%; p<0.001 by log rank test).

Conclusions The presence of neutralizing antibodies to the anti-TNF monoclonals is associated with lower trough levels of the drugs. Trough level of the anti-TNF biologics is useful for optimizing the clinical efficacy of the drugs in patients with inflammatory arthritis.

Disclosure of Interest None declared