Association Between Treatment with Infliximab and the Development of Tuberculosis
Braun, et al (N Engl J med 345(15):1098, 2001) reviewed the FDAs Adverse Event Reporting System (AERS) for all reports of tuberculosis with the treatment of infliximab for Crohns disease and rheumatoid arthritis (RA) from 1998 through May 29, 2001 to evaluate the timing and pattern of disease following initiation of infliximab therapy. It is estimated that approximately 147,000 patients worldwide have received infliximab therapy. Infliximab is a chimeric antibody directed against human TNF-a.
70 patients were reported to have developed tuberculosis beginning a median of 12 weeks (range, 1 to 52) after the initiation of infliximab therapy. The clinical presentation of these cases was atypical. 56% (40/70) of patients presented with extrapulmonary tuberculosis, and 24% (17/70) with disseminated disease, in contrast to 18% and 2%, respectively, of patients who have tuberculosis not associated with HIV or infliximab therapy (Cauthen, et al. Am Rev Respir Dis 51:141, 1990). 55 of the 70 patients reported had or were receiving one or more other immunosuppressive medications. 12 of the 70 patients subsequently died, and 4 of these deaths appear to have been directly related to TB.
Based on this report, the estimated rate of tuberculosis associated with infliximab in U.S. RA patients is 24.4 cases per 100,000 compared to a background rate of 6.2 cases per 100,000 among RA patients not treated with infliximab (Urbansky, et al. Arthritis Rheum. In press).
Conclusion: That these cases of TB are associated with infliximab therapy is suggested by the higher than background incidence, the temporal relationship between the two, and the unusual mode of presentation (high rate of disseminated disease). The rapidity of development of TB after initiation of infliximab suggests that this is re-activation of latent disease, rather than the development of new disease. Screening of patients with PPDs prior to initiation of infliximab is recommended.
Editorial Comment: The calculated incidence of infliximab-associated TB was based on a voluntary reporting system in which the true numerator and denominator for any adverse event is not truly known. However, an adverse event associated with a drug is more likely to be under-reported than over-reported. Interestingly, the number of cases of TB with etanercept, another TNF inhibitor was only 9 per 100,000 (using the same reporting system).
The association of infliximab with TB therefore appears to be real. TNF is a critical mediator in the containment and killing of mycobacteria and other intracellular pathogens, presumably by inducing apoptosis of alveolar macrophages infected with M. Tb. TNF also appears to be associated with the formation of caseating granulomas which contain latent infection with M. Tb. Chronic depletion of TNF by infliximab presumably removes these protective responses. That etanercept does not appear to carry the same risk for TB may be explained by the lower stability of its complexes with TNF, compared to infliximab.