There were 239 reports of NTM infection in patients who were receiving anti–TNF-α therapy. Most reports were for patients receiving infliximab (n = 174, 75%), followed by etanercept (n = 41, 17%), and adalimumab (n = 19, 8%). One case was reported in 1999 (patient used etanercept); numbers of reported infections among those using each product increased in 2001 and thereafter. Reported cases among those using each of the 3 drugs were highest in 2005 (). Of these reports, only 76 (32%) met either ATS/IDSA pulmonary disease criteria or our case definition for extrapulmonary disease. An additional 29 (12%) cases were judged to be probable cases, but the reports did not contain enough clinical or radiographic information to determine whether patients met ATS/IDSA NTM disease criteria. In other instances, the reports were either clearly not of cases of NTM disease (n = 27, 11%) or could not be determined (n = 95, 40%) because of a lack of microbiologic data, unclear reporting, or duplicate reports (n = 12, 5%). Of the 244 reports, 76 (31%) were from outside the United States (Europe, n = 40; Japan, n = 21; Canada, n = 4; Israel, n = 1; South Africa, n = 1; not specified, n = 9). Of patients with confirmed and probable cases (n = 105), a similar proportion (n = 35, 33%) were from outside the United States; most of these were from Europe (n = 15) or Japan (n = 12).
Figure 1 Case reports of nontuberculous mycobacteria in patients using antitumor necrosis factor-α (TNF-α) therapy, US Food and Drug Administration MedWatch database, 1999–2006. Cases are reported by each full year of data reporting for (more ...)
Of the 105 confirmed or probable cases, most were in women (n = 66, 65%), and the median age was 63 years (range 20–90 years). The anti–TNF-α agents reported for these patients included infliximab (n = 73, 69%), etanercept (n = 25, 24%), and adalimumab (n = 7, 7%). M. avium was the most common etiologic organism reported (n = 52, 49%), followed by rapidly growing mycobacteria (n = 20, 19%), and M. marinum (n = 8, 8%) (). Nine patients (9%) had died by the time their case was reported, and 64 (61%) had NTM adverse events that resulted in hospitalization. The most common underlying medical indication for anti–TNF-α therapy was rheumatoid arthritis (n = 73, 75%), followed by other inflammatory diseases (). Sixty-eight (65%) patients received concomitant prednisone, and 58 (55%) received methotrexate at the time of their report. Twenty-five (24%) patients reportedly had >1 of the following conditions: bronchiectasis (n = 5, 5%), chronic obstructive pulmonary disease (n = 11, 10%), diabetes mellitus (n = 5, 5%), and rheumatoid lung (n = 4, 4%). Median time between anti–TNF-α agent start date and infection diagnosis was available for only 68 (65%) of the patients. For adalimumab (n = 5), the interval was 18 weeks (range 4–94 weeks), for etanercept (n = 22) it was 35 weeks (range 0–288 weeks), and for infliximab (n = 41) it was 43 weeks (range 2–200 weeks).
Figure 2 Reported causes of 105 confirmed and probable nontuberculous mycobacteria (NTM) infections associated with antitumor necrosis factor-α agents, US Food and Drug Administration MedWatch database, 1999–2006. *Other species include Mycobacterium (more ...)
Reported diseases associated with anti–TNF-α therapy and therapy implicated for 105 cases of NTM disease, US Food and Drug Administration MedWatch database, 1999–2006*
The pulmonary region (n = 59, 56%) was the most frequently reported site of disease; the remainder of infections were extrapulmonary or disseminated (). Compared with patients with extrapulmonary NTM disease, patients with pulmonary NTM disease were more likely to have underlying rheumatoid arthritis (OR 3.6, 95% confidence interval [CI] 1.5–8.8, p<0.01) and more likely to be infected with M. avium (OR 11.0, 95% CI 4.4–27.9, p<0.01). Reported cases of pulmonary NTM disease were also more likely to be in female patients (OR 2.3, 95% CI 1.0–5.3, p = 0.04) (). After we adjusted for differences in sex in a stratified analysis, rheumatoid arthritis remained associated with pulmonary disease (adjusted OR 3.4, 95% CI 1.4–8.3, p = 0.01). There were no significant differences between the proportions of pulmonary disease, disseminated disease, death, or etiologic organisms reported for patients who used infliximab and those who used etanercept. Overall, infliximab users were more likely to be using methotrexate (OR 3.0, 95% CI 1.2–8.0, p = 0.02) than were etanercept users (63% vs. 36%). Prednisone use did not differ between the 2 groups; 50 (68%) of infliximab-treated patients and 14 (56%) of etanercept-treated patients were using prednisone at the time of their report (p = 0.19).
Sites of infection for 105 reported anti–TNF-α therapy–associated cases of NTM disease, US Food and Drug Administration MedWatch database, 1999–2006*
Characteristics of 105 pulmonary and nonpulmonary anti–TNF-α therapy–associated cases of NTM disease, US Food and Drug Administration MedWatch database, 1999–2006*
For reported NTM adverse events that did not meet either ATS/IDSA case criteria or our probable case designation, similar analyses were performed. These persons were similar to persons with cases that met confirmed or probable case definitions. They did not differ with regard to sex, underlying medical condition, anti–TNF-α drugs used, mortality rate, or the ratio of pulmonary to extrapulmonary disease manifestations. However, these patients were less likely to be using concomitant prednisone (65% for patients with cases vs. 43% for patients without cases; OR 2.5, 95% CI 1.4–4.2, p<0.01) or concomitant methotrexate (55% for patients with cases vs. 37% for patients without cases; OR 2.1, 95% CI 1.3–3.6, p<0.01).