cience Watch has covered some of the published clinical experience with "anti-tumor-necrosis-factor" (anti-TNF-alfa) agents, the most clinically well-researched of which are the drugs infliximab and etanercept. Recalcitrant rheumatoid arthritis and Crohn’s disease are the major, but by no means the only, conditions for which these agents hold promise. TNF is a protein that has an important role among the cast of players in immunity; interest in it goes far beyond the "tumor necrosis" that the name implies. If clinicians attempt to modify the body’s immune system in some way—which is what they are doing when they prescribe anti-TNF agents for patients whose diseases may have an element of autoimmunity in their causation—there may be a price to pay. With infliximab (paper #10), but possibly less so with etanercept, that price may be tuberculosis.

     Infliximab is not the treatment of first choice for rheumatoid arthritis; it is tried when all else seems to be failing. Dr. Joseph Keane and colleagues (#10) used reports to the U.S. Food and Drug Administration MedWatch program to study 70 cases of tuberculosis developing in patients who had been given infliximab. Tuberculosis is a bacterial infection that most patients and their families would assume to be a lung disease, but not for nothing was it classically known as "consumption," and two-thirds of these patients had manifestations outside the respiratory system. Four of the 12 deaths were probably related to the tuberculosis. The timing of the treatment in relation to the onset of the disease constituted the strongest evidence for cause and effect, and the most probable mechanism is reactivation of preexisting but dormant disease. There is animal evidence pointing to a role for TNF in protecting against Mycobacterium tuberculosis, so antagonizing TNF might well remove that protection.  

     Several other infections are also being recorded in patients taking this agent, examples being Pneumocystis carinii (a serious problem in patients with AIDS), Listeria monocytogenes (the microorganism behind periodic food scares over certain soft cheeses), and Histoplasma capsulatum. It is M tuberculosis, however, that is causing most concern, and patient-accessible websites already carry warnings. Keane’s paper (#10) attempts an estimate of the risk in numerical terms, and a recent review article from a Canadian group takes the calculations further (see M.A. Gardam, et al., Lancet Infect. Dis., 3[3]: 148-55, 2003). 

     Science Watch has previously highlighted a 1999 report of a randomized trial of infliximab in severe rheumatoid arthritis (R. Maini, et al., Lancet, 354[9194]: 1932-9, 1999), and it was in that study that the first case of tuberculosis possibly associated with the drug was recorded. So Science Watch asked the lead author on that paper, Professor R.N. Maini from the Kennedy Institute of Rheumatology in London , U.K. , what his take is on the later report (#10). He notes that "Experimental evidence and the mechanism of action of anti-TNF agents demonstrates a major effect on the recruitment of lymphocytes and monocytes to the site of inflammation. Hence maintenance of a granuloma to contain latent tuberculosis could well be compromised by anti-TNF therapy." In clinical practice patients must be screened for latent tuberculosis before infliximab is given.