TNFα antagonist therapy does not increase the Epstein-Barr virus burden in patients with rheumatoid arthritis or ankylosing spondylitis.

Objective: The risk of non-Hodgkin lymphoma is increased in rheumatoid arthritis (RA) but not in ankylosing spondylitis (AS). In RA, the degree of inflammation is closely associated with the lymphoma risk. Whether immunosuppressants such as methotrexate and TNFα antagonists affect the lymphoma risk in RA is unclear. The Epstein-Barr virus (EBV) may contribute to the pathogenesis of RA and may be involved in the development of lymphoma in patients taking methotrexate and/or TNFα antagonists, although these points remain debated. EBV load monitoring during immunosuppressive treatment may predict the occurrence of EBV-related lymphoma. Here, our objective was to prospectively measure the EBV load in patients receiving TNFα antagonists for RA or AS.

Methods: We prospectively studied patients with RA or AS before and after TNFα antagonist therapy initiation. The EBV load was measured in blood samples using the EBV R-gene Quantification Kit. Disease activity at the time of blood sampling was evaluated by determining the Disease Activity Score 28 (DAS28) in RA patients and the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) in AS patients.

Results: We included 46 patients with RA (82.6% women; mean age, 52.7 ± 11.3 years) and 27 with AS (men, 81.5%; mean age, 45.1 ± 12.7 years). In the RA group, the EBV load was measured at baseline and 9.72 ± 5.7 months later. The baseline EBV load was undetectable in 33 (70.2%) patients; mean EBV load in the 13 remaining patients was 9389 copies/ml (3.47 log₁₀ ± 0.45). Baseline EBV load did not correlate with disease activity (DAS28). At the follow-up assay, the EBV load became positive in five patients and increased significantly in one patient (four patients on etanercept, one on adalimumab, and one on infliximab); it became negative in six patients (five on adalimumab and one on etanercept) and showed non-significant changes in six patients. Mean EBV load in patients positive at follow-up was 3.63 ± 0.52 log₁₀ copies/ml. Mean DAS28 was 4.78 ± 1.1 at baseline and 2.94 ± 1.24 at follow-up. At follow-up, a good EULAR response was noted in 33 (71.7%) patients and a moderate EULAR response in seven (15.2%) patients. In the AS group, the baseline EBV load measurement occurred after 12.9 ± 10.6 months. Baseline EBV load was undetectable in 25 (92.6%) patients; mean load in the remaining two patients was 4.15 ± 0.46 log₁₀ copies/ml. At follow-up, the EBV load became positive in two patients (one on adalimumab and one on infliximab) and became negative in one patient (on adalimumab). Mean load in positive patients was 3.33 ± 0.24 log₁₀ copies/ml. Mean BASDAI was 55.1 ± 16.2 at baseline and 17.88 ± 18.62 at follow-up. A positive EBV load was significantly more common in the RA group than in the AS group (P = 0.039). EBV load changes did not differ significantly between the RA and AS groups or across TNFα antagonists. No cases of lymphoma were recorded.

Conclusion: Introducing TNFα antagonist therapy does not affect the EBV load in patients with RA or AS. EBV load monitoring is probably unnecessary in patients given TNFα antagonists for RA or AS.