SNOWMASS, COLO. – Rheumatoid arthritis patients on tumor necrosis factor inhibitors are at markedly increased risk for both tuberculosis and nontuberculous mycobacterial lung disease, as highlighted in data not yet published from Kaiser Permanente of Northern California that was discussed by investigator Dr. Kevin L. Winthrop at the symposium.

The crude incidence rate of nontuberculous mycobacterial disease in this 3.1-million-member health plan during the study years of 2000-2008 was 4.1 cases/100,000 person-years. The risk rose with age such that among plan members aged 50 years or older the rate was 11.8 cases/100,000 person-years. Plan members with rheumatoid arthritis (RA) who’d never been on a tumor necrosis factor (TNF) inhibitor had a moderately higher rate of 19.2 cases/100,000 person-years, probably because of their use of prednisone. But among 8,418 RA patients on TNF inhibitor therapy, the incidence of nontuberculous mycobacterial pulmonary disease shot up to 112 cases/100,000 person-years.

The tuberculosis incidence followed a similar pattern: 2.8 cases/100,000 person-years among the general Kaiser membership, 5.2 in those aged 50 years or older, 8.7 in RA patients never exposed to a TNF inhibitor, jumping to 56 cases/100,000 person-years among RA patients on an anti-TNF biologic, according to Dr. Winthrop.

In light of these data, physicians need to be on the lookout for nontuberculous mycobacterial pulmonary disease arising in RA patients using a TNF inhibitor.

"You will have patients with this. It’s best to diagnose them early, if possible, and get them off their biologic and also limit or discontinue prednisone," said Dr. Winthrop, an infectious diseases specialist at Oregon Health and Science University, Portland.

It’s his clinical impression, as well as that of other physicians participating in the Infectious Diseases Society of America’s Emerging Infections Network, that people who develop nontuberculous mycobacterial lung disease while on a TNF inhibitor tend to experience more rapid progression of their lung disease, he said at the meeting.

He and his coinvestigators at Kaiser also examined the pulmonary disease rates associated with individual TNF inhibitors. The nontuberculous mycobacterial lung disease rate in patients on etanercept was 35 cases/100,000 person-years of exposure, significantly less than the 116 cases/100,000 person-years with infliximab or 122 with adalimumab.

Similarly, the tuberculosis rate was lowest with etanercept at 17 cases/100,000 person-years as compared to 83 with infliximab and 61 with adalimumab.

The Kaiser experience confirms a 2010 report from the British Society for Rheumatology biologic registry that provided the first solid epidemiologic data showing that TNF inhibitors carry an increased tuberculosis risk. In the U.K. study, etanercept use was associated with a tuberculosis incidence of 39 cases per 100,000 person-years, significantly lower than the 136 cases per 100,000 person-years with infliximab or 144 with adalimumab. In contrast, the tuberculosis rate among more than 3,200 RA patients on a conventional disease-modifying anti-rheumatic drug was zero. The background tuberculosis incidence in the United Kingdom during that time period was about 12 cases/100,000 person-years (Ann. Rheum. Dis. 2010;69:522-8).

"I’m convinced that the monoclonal antibody TNF inhibitors [infliximab and adalimumab] cause more tuberculosis than [does] etanercept. I’m not convinced I know why," Dr. Winthrop admitted.

Numerous potential mechanisms have been floated to explain this differential effect. The two he finds most plausible are that etanercept is less able to penetrate tuberculosis granulomas than are the monoclonal antibody TNF inhibitors, as shown in a mouse model, and the possibility – as yet unproven – that etanercept might also cause less downregulation of CD8 cells producing the antimicrobial peptides perforin and granulysin, which are directed against Mycobacterium tuberculosis.

Dr. Winthrop said the epidemiology of nontuberculous mycobacterial pulmonary disease is changing. Decades ago, it was viewed as a disease of elderly men. As the incidence has climbed during the past 2 decades, however, the disease has come to be recognized as mainly one of postmenopausal women, typically with no history of underlying lung disease or smoking. The phenotype is one of an elderly woman who is tall, slender, and underweight, often with mitral valve prolapse, scoliosis, or pectus defects.

In a large study conducted at four geographically diverse large health plans, the annual prevalence of nontuberculous mycobacterial lung disease among persons age 60 years or older rose from 19.6 cases/100,000 in 1994-1996 to 26.7 cases/100,000 person-years in 2004-2006, a rate two- to threefold greater than the prevalence of tuberculosis at those sites during 2004-2006 (Am. J. Respir. Crit. Care Med. 2010;182:970-6).

Nontuberculous mycobacteria are ubiquitous in tap water and soil. Unlike tuberculosis, which is spread from person to person by coughing, nontuberculous mycobacterial infections are acquired directly from the environment.

Dr. Winthrop said he is reluctant to recommend resuming biologic therapy after a RA patient has been treated for nontuberculous mycobacterial lung disease or coccidioidomycosis.

Dr. Winthrop reported having received consultant fees from Abbott, Amgen, and Pfizer as well as research funding from Pfizer.

Anticipate Mycobacterial Lung Disease in Anti-TNF Users