Brief Communication: Characteristics of Spontaneous Cases of Tuberculosis Associated with Infliximab FREE

—The Editors

Infliximab (Remicade, Centocor, Malvern, Pennsylvania) is a chimeric monoclonal antibody to tumor necrosis factor-α (TNF-α) that is approved by the U.S. Food and Drug Administration (FDA) for treating Crohn disease, rheumatoid arthritis, ulcerative colitis, ankylosing spondylitis, psoriatic arthritis, psoriasis, and pediatric Crohn disease (1). Tumor necrosis factor-α has a central role in both the host immune response to Mycobacterium tuberculosis infection and the immunopathology of tuberculosis (2). In addition, TNF-α production is required for the formation of granulomas, which help to sequester mycobacteria and prevent their dissemination (3). Therefore, it was not surprising that the FDA received adverse event reports of tuberculosis in association with infliximab treatment after the initial marketing of infliximab in the United States in September 1998. These reports were summarized in an article published in October 2001 (4). Coincidentally, a boxed warning was added to the infliximab package insert in October 2001, advising physicians to screen patients for tuberculosis before treatment with infliximab therapy, pretreat those with latent disease, and monitor patients for signs and symptoms of tuberculosis during treatment with infliximab. It should be noted that approved labeling for the 2 other commercially available TNF-α antagonists in the United States, etanercept (Enbrel, Immunex Corporation, Thousand Oaks, California) and adalimumab (Humira, Abbott Laboratories, North Chicago, Illinois), also include warnings about tuberculosis.

We examined spontaneous adverse event reports of tuberculosis associated with infliximab that were submitted to the FDA from 1 November 2001 through 30 May 2006, after the boxed warning was added to the drug labeling, to assess the continuing safety concern of infliximab-associated tuberculosis. We analyzed the reports to determine whether the tuberculosis cases were due to failure of prescribers to test for latent tuberculosis, as mandated in the prescribing information, or (if prescribers are performing tuberculosis screening) due to false-negative purified protein derivative results. We also examined whether patients with latent tuberculosis were being treated before starting infliximab therapy and, if so, whether those patients subsequently developed tuberculosis. In addition, we analyzed patient risk factors associated with the development of tuberculosis during treatment with infliximab.

Health care professionals, consumers, drug manufacturers, and others submit reports of adverse events associated with FDA-licensed drug products through interaction with drug manufacturer representatives or directly via the FDA's MedWatch program (5). Because reporting is voluntary for everyone except manufacturers, the proportion of all incident adverse events attributable to a drug product that are actually reported to the FDA is unknown (5 - 6). Adverse drug event reports (also known as MedWatch reports) submitted to the FDA are maintained and retrieved through the FDA's Adverse Event Reporting System (AERS) database (7). We used the Medical Dictionary for Drug Regulatory Affairs grouping term tuberculous infection to retrieve adverse event reports of tuberculosis from this database. We included reports of cases of active tuberculosis among patients receiving infliximab in the United States with received dates ranging from 1 November 2001 through 30 May 2006.

After reviewing each of the 278 reports of infliximab-associated tuberculosis reported to the FDA within the specified time interval, we included 130 reports within the case series and excluded 148. The most frequent reasons for exclusion included duplicate reports, reports of latent disease, foreign (non-U.S.) reports, hearsay reports, reports from a clinical trial or registry, nontuberculous mycobacterial infection, and reports from patients more than 1 year after cessation of infliximab therapy.

By year, the case series contains 9 reports from 2001, 35 from 2002, 30 from 2003, 24 from 2004, 21 from 2005, and 11 from 2006. Table 1 shows the demographic and clinical characteristics of the patients. Four reports of tuberculosis were in patients treated for psoriasis (1 patient was concurrently being treated for rheumatoid arthritis). The median time to onset from initiation of infliximab therapy to diagnosis of tuberculosis was 10 months, with a range of 1 to 72 months. Twenty-two cases reported tuberculosis onset within 3 months, and 40 cases reported time to tuberculosis onset as greater than 1 year. Approximately one third (37%) of cases reported tuberculosis restricted to the lungs; 31 (23%) cases reported patients with disseminated tuberculosis; and 28 reported extrapulmonary tuberculosis. Extrapulmonary tuberculosis included peritoneal disease (n = 7), lymph node disease (n = 4), bone or joint disease (n = 4), enteric disease (n = 4), meningeal disease (n = 4), hepatic disease (n = 2), pericardial disease (n = 2), and pelvic disease (n = 1). Sixty-seven patients were hospitalized, and 19 died.

The 19 patients who died were nearly equally distributed in sex and ranged from 26 to 78 years of age, with a median age of 61 years. The most commonly reported indication was for the treatment of rheumatoid arthritis (68%). Infection was identified most frequently as the cause of death (n = 10), including 6 cases in which tuberculosis was the specified cause of death. Most (89%) of the fatal cases reported concomitant risk factors for tuberculosis. Fourteen patients were reported to have received 1 or more immunosuppressive medications concomitantly. Risk factors other than immunosuppressant use were present in 9 cases. These factors included having been born, living, or traveling in an area with an increased tuberculosis rate; exposure to a person with tuberculosis; gastrointestinal resection; and occupational exposure.

Most patients seemed to have at least 1 risk factor for tuberculosis other than infliximab use, including 52 patients with at least 1 risk factor other than immunosuppressant use. The most commonly reported risk factor other than immunosuppressant use (68%) was a history of latent or active tuberculosis (n = 33 [25%]). Seventeen (55%) of these 33 cases reported previous treatment for tuberculosis. Twenty-five (20%) cases described patients who were born or had spent extensive time in an area where tuberculosis is endemic.

The revised infliximab label recommends that patients receive a tuberculin skin test before starting infliximab therapy. We conducted an analysis of the skin test results from the subset of 67 cases who started infliximab therapy after the addition of the boxed warning in October 2001 (Table 2). By report, 47 (70%) patients received a tuberculin skin test before initiation of therapy. Of these patients, 34 had a negative initial result (insufficient data were provided on the actual induration size). Two of the 6 patients who did not receive a tuberculin skin test before starting infliximab therapy listed the reason as “prior bacille Calmette–Guérin vaccination.” Of these 67 cases, 12 had a history of latent or active tuberculosis. As described in the case reports, 8 of these 12 individuals were treated for tuberculosis before starting infliximab therapy.

Eleven patients had received a bacille Calmette–Guérin vaccine. Of these, 6 received a tuberculin skin test before starting infliximab therapy; 2 had negative results and 4 had positive results. Two patients did not receive a tuberculin skin test because of their previous bacille Calmette–Guérin vaccination. Whether the remaining 3 patients were tested before starting infliximab therapy is unknown.

We also analyzed the subset of 67 patients for the results of prescreening chest radiography performed before infliximab therapy. Twenty-three patients had had radiography done before starting infliximab therapy. Of those patients, 22 had a normal chest radiograph and 1 result was unknown. Fourteen of the patients with a normal chest radiograph developed pulmonary tuberculosis, 5 developed extrapulmonary tuberculosis, and 3 developed disseminated tuberculosis.

The infliximab label carries a prominent warning about the risk for tuberculosis. Nineteen (15%) of the 130 cases of infliximab-associated tuberculosis that we studied were associated with a fatal outcome. It is important to note that almost half (45%) of the cases of infliximab-associated tuberculosis reported to the FDA since the revision of the product labeling in 2001 presented with disseminated (23%) or extrapulmonary (22%) disease. Clinicians caring for patients who are receiving infliximab therapy should be aware of the clinical manifestations of disseminated or extrapulmonary tuberculosis. In addition, these data suggest that most of the reported patients who were treated with infliximab after the labeling change received a tuberculin skin test before starting infliximab therapy. Patients who had had a negative result developed active tuberculosis after starting therapy, as have patients treated for latent tuberculosis or tuberculosis—suggesting that sterilization of the M. tuberculosis bacterium may not have been achieved in some cases.

These data also suggest that patients with demographic risk factors, such as having been born, living, or traveling in a country with higher background tuberculosis rates than those of the United States, may be at an increased risk for active tuberculosis while receiving infliximab therapy. We found reports of patients with previous bacille Calmette–Guérin vaccination who developed active tuberculosis after initiation of infliximab therapy, both with and without screening for latent tuberculosis before infliximab therapy.

Clinicians should be aware of the risk for tuberculosis and the need to appropriately screen patients for tuberculosis risk factors before starting infliximab therapy, as outlined in the current drug labeling. In this highly immunosuppressed patient population, false-negative tuberculin skin test results can occur. Maximizing the sensitivity of the tuberculin skin test by using a cutoff induration size of 5 mm and taking a thorough patient history with regard to tuberculosis risk factors before and during infliximab therapy are warranted (8). Physicians should also monitor such patients for signs and symptoms of extrapulmonary tuberculosis. Because patients may develop active tuberculosis even when their prescreening tuberculin skin test result is negative, physicians need to evaluate for tuberculosis risk factors before and during therapy. Communication of these risks to physicians and patients is important to help mitigate the risk for active tuberculosis in patients treated with infliximab.