Hotline

FDA Advisory Committee Reviews Safety of TNF Inhibitors

On August 17, 2001 the FDA convened a meeting of the Arthritis Advisory Committee (AAC) and the sponsors of etanercept (Immunex) and infliximab (Centocor) to review and update past and current safety observations related to the use of TNF inhibitors. Representatives of the Center for Biologics Evaluation and Research (CBER) reported to the committee on their analysis of adverse events reported to the FDA through Medwatch and the Adverse Event Reporting System (AERS). Representatives from Immunex and Centocor addressed the AAC regarding their safety database as well as their current and future pharmacosurveillance efforts to educate physicians and patients regarding these matters. Serious and unusual adverse events potentially related to TNF inhibitors were reviewed. The table below summarizes available data from the meeting.

As of July 2001 over 270,000 patients have been exposed to etanercept or infliximab worldwide. Whereas infliximab is marketed for use in Europe (representing 18% of worldwide use), etanercept is available on a more limited basis in Europe (accounting for ~6% of its worldwide use). Whereas > 90% of etanercept use has been in RA, most of the worldwide use of infliximab has been in Crohn’s disease more so than RA. Foremost in the discussion were concerns raised by reports of tuberculosis (M.Tb) and other opportunistic infections infrequently observed in patients taking TNF inhibitors. M.Tb has been reported in 82 infliximab-treated and 11 etanercept-treated patients worldwide. Importantly, 80% of the infliximab cases were from outside of the US, predominantly from Europe and Norway. M.Tb cases with infliximab occurred during treatment for RA (67.5%), Crohn’s disease (26%) and other indications (4.5%). In comparison, 11 reports of M.Tb and 8 cases of atypical mycobacterial infections were seen in etanercept treated patients. Infliximab-related cases were discovered in the first 6 months of use, suggesting reactivation of latent M.Tb, with nearly 80% of these occurring outside the USA. Nearly one-third of cases had an extrapulmonary or miliary presentation.

Risk factors for M.Tb (and other opportunistic infections) appear to include current and past residence in endemic regions (e.g., Europe for M.Tb, Ohio River Valley for histoplasmosis, San Joaquin Valley for coccidioidomycosis) and chronic prednisone (>15 mg/day) use. The American Thoracic Society (ATS) and Centers for Disease Control and Prevention (CDC) also include recent immigrants (from high prevalence countries), IV drug users, those exposed to M.Tb, lab personnel, and patients with silicosis, diabetes mellitus, chronic renal failure, leukemia, lymphoma, and HIV (+) as high risk populations. However, most of the RA patients (Table 1) had no identifiable risk factor other than treatment with a TNF inhibitor. A majority of RA patients in the table below were receiving prednisone and other DMARDs at the time of reported infection. Although chronic prednisone use is a risk factor for serious or opportunistic infections, it is unknown whether methotrexate or other DMARDs significantly influences the incidence of opportunistic infections in RA.

Data from preclinical animal models indicate that TNF, nitric oxide and other proinflammatory cytokines play an important and specific role in the containment and isolation of viable tubercle bacilli, presumably by promoting macrophage activity and granuloma formation. Likewise, TNF knockout mice are more susceptible to infection with intracellular pathogens such as candida, listeria and M.Tb. Thus, the inhibition of TNF by these biologic agents may facilitate reactivation of latent M.Tb in some patients.

The package insert for infliximab has been revised by Centocor warning that M.Tb and other opportunistic infections have been observed with infliximab use, and advising physicians to employ PPD skin testing in patients who will receive or are receiving infliximab. Studies have shown no support for chest X-ray (CXR) or skin controls (e.g., Mumps, candida) as screening measures for latent M.Tb. Discussants at the meeting considered whether these findings indicated a “class effect” and if these recommendations would apply to all TNF inhibitors. This issue awaits further study for each drug. It is unclear if skin testing should be use in those on etanercept. However, some rheumatologists (who are concerned about a class effect) may use skin testing in their etanercept-treated patients until further studies become available. Immunex recommends that physicians follow ATS/CDC guidelines on the use of targeted skin testing¹.

Other opportunistic infections were reported with infliximab and etanercept (Table 1), including histoplasmosis (9 vs. 1, respectively), listeriosis (11 vs. 1), pneumocystis (12 vs. 5) aspergillosis (6 vs. 2), and atypical mycobacteria (0 vs. 8). As no reliable skin testing exists for these infections, clinicians must closely monitor patients for symptoms or signs of opportunistic infections, especially in endemic areas. Patients found to have a +PPD should be further evaluated for active infection (e.g., CXR, sputum AFB) and should be treated according to ATS/CDC recommendations⁽¹⁾.

Other safety issues were raised at the meeting. The number of cases of demyelinating disorders, multiple sclerosis, optic neuritis, pancytopenia, and aplastic anemia are presented below (Table1). New information regarding reports of patients with seizures, colonic perforations, lupus-like manifestations and lymphoma in patients receiving TNF inhibitors was also presented. It was universally felt that prospective observational programs are needed to further confirm the long-term safety of TNF inhibitors. Both sponsors detailed their comprehensive pharmacosurveillance programs (>15,000 patients to be observed prospectively) to address these specific issues.

Rheumatologists should carefully monitor patients on these agents using sponsor recommended guidelines. Patients taking TNF inhibitors who exhibit serious adverse events, including M.Tb, opportunistic infection or other unusual or life-threatening clinical disorders, should be reported to the FDA through the sponsor or by filling out forms available at the Medwatch Web site (http://www.Medwatch.com)

Infectious agent or Adverse Event	Infliximab (n~170,000)^a	Etanercept (n~104,000)^a	Historic Population Incidence Rate^b
Mean Age (yrs)	53	56	-
M. tuberculosis	84^c	11	USA 8.2/100,000 pt-yrs)^d RA 6/100,000^e
Atypical mycobacterium^f	0^c	8	NA
Histoplasmosis	9	1	NA
Listeria monocytogenes	11	1	NA
Pneumocystis carinii	12	5	NA
Aspergillosis	6	2	NA
Candiasis	7	3	NA
Cryptococcosis	2	3	NA
Coccidioidomycosis	2	0	NA
Pancytopenia	15	12	NA
Aplastic anemia	0	4	RA 5.7-8.2/100,000 pt-yrs.
Multiple sclerosis: Total	6	14	NA
Multiple sclerosis: New dx	3	6	4/100,000 pt-yrs
Optic neuritis	4	3	5/100,000 pt-yrs
Seizures	29	26	35/100,000 pt-yrs
Lupus-like disease	4	4	NA
Colonic perforations	NA	13	NA
Lymphoma	10	18	NA

^a) Number exposed to drug worldwide;

^b) Reference incident rates are historic and derived from the literature. Comparison with individual drug reporting rates is flawed for many reasons, including an unknown magnitude of underreporting of adverse events;

^c)FDA presentation reported 92 cases of M.Tb with infliximab as of 8/17/01 and 3 cases of atypical Tb have since been reported;

^d) US age adjusted incidence rate expressed per 100,000 patient-year;

^e) Wolfe F: to be presented 2001 ACR annual meeting, San Francisco;

^f) includes 6 cases of M. avium intracellulare and 1 each of M. kansansii and M. marinarum

September, 2001
John J. Cush (2), MD, Eric L. Matteson, MD, MPH
Hotline Editors

¹ ATS/CDC Guidelines (Am J Respiratory Crit Care Med 161:S221-S227, 2000)

²Dr. Cush is a paid consultant, lecturer and researcher for Immunex and Centocor

(Acknowledgements: the authors want to thank the instructive input of Drs. Vibeke Strand, Arthur Kavanaugh, Jeffrey Siegel (FDA/CBER) and officials at Centocor and Immunex in the preparation of this report)