SNOWMASS, COLO. – The risk of herpes zoster isn’t raised in rheumatoid arthritis patients who are on a tumor necrosis factor inhibitor, compared with those on nonbiologic disease-modifying antirheumatic drugs, according to a preliminary analysis from by far the largest study to examine the question.
Findings from the same study also suggested that TNF–inhibiting therapy may actually protect against zoster in patients with psoriasis, psoriatic arthritis, ankylosing spondylitis, or inflammatory bowel disease.
Dr. Kevin L. Winthrop
Data from two previous, large, population-based studies looking at the risk of herpes zoster in RA patients on anti-TNF biologics reveal discordant conclusions. These conflicting findings prompted Dr. Kevin L. Winthrop and his coinvestigators in the SABER (Safety Assessment of Biologic Therapy) collaboration to examine the issue in their very large data set.
SABER is a retrospective cohort study involving four large U.S. automated health databases. The as-yet-unpublished SABER herpes zoster analysis included more than 35,000 patients with diseases for which TNF inhibitors are indicated.
During 22,215 person-years of follow-up, the crude rate of herpes zoster among RA patients who were new first-time users of an anti-TNF agent was 12.1 cases per 1,000 person-years, which wasn’t significantly different from the 12.8 per 1,000 rate during 7,165 person-years of follow-up in the control group, which comprised RA patients who were failing on methotrexate and adding another conventional DMARD, usually hydroxychloroquine (Plaquenil).
There were no significant differences in zoster rates among the various anti-TNF agents, Dr. Winthrop said at the symposium, which was sponsored by the American College of Rheumatology.
Intriguingly, going on an anti-TNF agent actually appeared to protect against herpes zoster in SABER participants with psoriasis, psoriatic arthritis, or ankylosing spondylitis. During 4,106 person-years of follow-up, their crude rate of zoster was 4.4 cases per 1,000 person-years, an adjusted 36% lower risk than that in patients on a nonbiologic DMARD, whose rate was 6.8 cases per 1,000 person-years during 3,950 person-years of follow-up, according to Dr. Winthrop, an infectious diseases specialist at Oregon Health and Science University, Portland.
Moreover, new users of infliximab or adalimumab for inflammatory bowel disease had an adjusted 20% reduction in risk of zoster, compared with those on azathioprine or 6-mercaptopurine, he added.
An earlier retrospective cohort study involving 20,357 RA patients in the Veterans Affairs health care system found that patients on adalimumab had an adjusted 47% lower risk of zoster, and those on etanercept had a 38% reduction in risk, compared with patients on conventional DMARDs. Patients on infliximab had a 30% increased risk; however, this trend didn’t reach statistical significance (Clin. Infect. Dis. 2009;48:1364-71).
In sharp contrast, investigators from the prospective German RABBIT registry involving 5,040 RA patients reported that adalimumab was associated with a 3-fold increased risk of herpes zoster compared to nonbiologic DMARD therapy. Infliximab had an adjusted 2.4-fold increased risk, which wasn’t statistically significant. Etanercept wasn’t associated with increased risk (JAMA 2009;301:737-44).
The explanation for the discordant findings in RABBIT and the VA study is unclear. The studies did use somewhat different methodologies. Regardless, it’s worth noting that the number of subjects included in the new SABER analysis was larger than for both of the earlier studies put together.
"Look how robust this is," Dr. Winthrop observed. "We have good, good power. Our interpretation is there is no increased risk of zoster with anti-TNF drugs."
Nonetheless, this is a preliminary analysis. The SABER investigators plan to look next at what happened to prednisone dosing when the RA patients went on anti-TNF therapy or an additional nonbiologic DMARD, something they can do with this cohort but couldn’t do in an earlier SABER publication. Dr. Winthrop said he suspects that the prednisone dose was more likely to be reduced or eliminated in the TNF-inhibitor users, and that this might account for the finding that anti-TNF therapy didn’t increase zoster risk. A half-dozen studies have shown that prednisone therapy is associated with an increased risk of herpes zoster.
Dr. Winthrop said it’s his impression that patients who develop herpes zoster while on anti-TNF therapy don’t have a higher rate of disseminated disease or worse outcomes than do those who are not. In the SABER analysis, the rate of hospitalization for zoster was similarly very low in the biologic and nonbiologic treatment groups.
In response to audience questions, he said he believes it makes sense to give herpes zoster vaccine to RA patients starting at 50 years of age. RA per se is a risk factor for zoster, as is prednisone therapy. There is no evidence of any safety issues in vaccinating patients who take any of the standard DMARDs or anti-TNF agents, although there are no definitive data on that score. It’s possible that pharmacologic immunosuppression blunts the vaccine’s efficacy, but that hasn’t been studied.