Background Patients (pts) on tumor necrosis factor inhibitors (TNFi) are at risk for infections with Mycobacerium tuberculosis (MTB) and non-tuberculosis bacteria (NTM). Prior to TNFi therapy, pts are screened for latent TB infection (LTBI) with tuberculin skin test (TST) or Quantiferon (QFN) assay (Cellestis) and chest x-ray (CXR). High risk pts as defined by Centers for Disease Control, on long term TNFi are still at risk for exposure and development of new infections.1 Guidelines on surveillance in pts on TNFi are currently lacking.
Objectives To best identify mycobacterial infection in TNFi- naïve and exposed pts using TST, QFN, or both. Additionally, to differentiate between latent or pre-symptomatic disseminated infection in new converters on TNFi using multi-site cultures (sputum, urine, and stool).
Methods We evaluated 580 pts treated with TNFi between 11/1/2000 and 10/26/2011 at a single county institution. Screening TST and/or QFN, as well as CXR were obtained at baseline. Negative pts were serially re-tested while on TNFi. Those with new positive TST (>5mm induration) and/or QFN underwent multi-site cultures. Asymptomatic pts with negative cultures were treated as LTBI with Isoniazid for 9 months; those with positive cultures were referred for evaluation and treatment. Non-parametric and Fisher’s exact tests evaluated differences between groups.
Results LTBI was prevalent at baseline; 169 of 580 (29%) pts were positive by either TST and/or QFN (figure 1). In TNFi-naïve pts dual testing uncovered a higher LTBI rate (38.3%) compared to TST (24.5%) or QFN alone (32%, p-trend=0.04). New conversions while on TNFi remained high; 51 of 440 (11.6%) serial re-screens were positive by either test. Dual testing uncovered more new conversions (15.2%) compared to TST (8.8%) or QFN alone (8.8%, p-trend=0.1). Of 51 new asymptomatic converters on TNFi, 50 completed multisite cultures; 10 (20%) had positive cultures, 1 (2%) for MTB and 9 (18%) for NTM. Of 169 TNFi-naïve pts with positivity for either test, 39 completed cultures; 4 (10%) were positive, 1 (2.6%) for MTB and 3 (7.4%) for NTM.
Conclusions LTBI prevalence is significant at baseline and its incidence remains so in high-risk pts exposed to TNFi. Dual testing with TST and QFN captures a higher rate of mycobacterial infections especially in naïve pts. A sizeable proportion of asymptomatic new converters on TNFi have positive cultures raising concern for dissemination. We therefore propose that both TNFi-naïve and exposed, high-risk pts undergo dual baseline and serial testing. New converters by either TST or QFN should be further evaluated with multi-site cultures for possible dissemination.
CDC MMWR 2005; 54 (RR-17): 1-141
Disclosure of Interest None Declared