Background TNF-α-blocking therapy is a successful treatment strategy for rheumatoid arthritis (RA), but is associated with reduced resistance to tuberculosis. We have developed a vaccine against mouse TNF-α (TNF-KLH) which protects mice from collagen-induced arthritis (CIA) and collagen-antibodies-induced arthritis (CAIA). We aimed at modelizing the infectious risk/benefit ratio of TNF-α neutralisation by comparing different strategies of TNF-α targeting in experimental Mycobacterium tuberculosis infection.
Materials and methods 4 groups (TNF-KLH, etanercept, KLH, PBS) of 10 C57BL/6 mice were compared with TNF deficient mice (TNF -/-). Vaccines were emulsified in IFA before i.m. administration (days −44, –31, −17 and −4). Other treatments were given from D0. All groups were infected at day 0 by 104 cfu of Mycobacterium tuberculosis (H37Rv virulent strain). Mouse groups were divided in two arms: one was euthanized 28 days post-infection, the lasting mice were euthanized 56 days post-infection. Bacterial burden was evaluated in lungs. Cellular infiltration was studied by immunohistological analysis.
Results A sustained anti-TNF-α antibody production was obtained in vaccinated mice. At day 28 of Mycobacteria infection, bacterial burden, organ weights, neutrophils and B cells infiltration were similar between TNF-KLH, KLH and PBS groups. At this time point, histological analysis showed higher surface of granuloma in liver of etanercept and TNF-/- groups than TNF-KLH group. At day 56 post infection, etanercept group presented higher surface of liver and lung granuloma than TNF-KLH and KLH groups. These observations were in accordance with a progression of the disease in etanercept group, whereas such effect was not retrieved in TNF-KLH and immunocompetent groups (KLH, PBS).
Conclusion Our results indicate that anti-TNF-α vaccination could protect mice from inflammatory arthritis without deeply altering host immunity against infection, suggesting the existence of two distinct pathophysiologic thresholds for TNF-α inhibition, one for arthritis improvement, the other for infection facilitation. These thresholds might be dependend on the TNF inhibition strategy.