News

Serious Infection Risk Similar for Biologics, Nonbiologics

View on the News

A New View on TNF Safety

Previous studies of anti–tumor necrosis (TNF) factor therapies have found an initially high rate of serious infections that declined with prolonged use of biologic therapies. Subsequent studies have suggested that the pattern is largely due to changes in glucocorticoid use and initial exclusion of high-risk patient who have their infections early and discontinue treatment, leaving a cohort of patients at low risk for infections, Will Dixon, Ph.D., and Dr. David Felson wrote in an accompanying editorial (JAMA 2011 Nov. 6; doi:10.1001/jama.2011.1705).

Unlike these studies, which included cohorts from users of disease-modifying antirheumatic drugs, Dr. Grijalva’s study selected new users of nonbiologic therapy who did not benefit from methotrexate in the rheumatoid arthritis cohort. Thus, the comparative risks of starting one treatment vs another can be examined.

"In the study ... the absolute rate of infection was much higher than in other studies of patients using disease-modifying anti-rheumatic drugs, perhaps reinforcing that the difference may be driven by differences in the comparator. This is possible because the comparison group, like the treatment group in this study, consisted of underserved, vulnerable patients – a population typically excluded from clinical trials."

The authors balanced the measured variables, but they could only match items that were measured. There was no way to know if anti-TNF therapy was avoided in patients because of prior infection or frailty.

The study also revealed a high rate of attrition. "Within the first 60 days of follow-up, approximately 40% of the comparison cohort was censored (i.e. did not continue to take their drugs) vs. approximately 15% of the anti-TNF group. By 120 days of follow-up, when other studies have suggested the risk of infection in TNF inhibitor users is high, more than 50% of users of comparator drugs and more than 30% of those taking TNF inhibitors were lost to follow-up."

This extensive loss is surprising and suggests an uncommon pattern of use. The cohorts were balanced for measured variables at the start of the study, but that balance may have quickly disappeared.

"Recent studies suggest that older patients who initiate anti-TNF therapy might quickly stop its use. If patients stopping TNF inhibitor treatment were at relatively high risk of infection, the risk of serious infection in drug users would be underestimated. However, it is impossible to tell the extent or direction of this bias, because no information is available on patients who were lost to follow-up."

The findings may prompt a reevaluation of anti-TNF safety.

Dr. Felson is chief of the Multidisciplinary Clinical Research Center at the Boston University School of Medicine. Dr. Dixon is a pharmacoepidemiologist at the University of Manchester, England. Neither reported any financial conflicts.


 

FROM THE JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION

During the first year after starting a new therapy, patients with autoimmune diseases appear to experience a similar rate of serious infections irrespective of whether the new drug is a biologic or a nonbiologic agent.

Infections requiring hospitalizations were nearly identical, regardless of the initial therapy, in a large retrospective cohort study of patients with rheumatoid arthritis (RA), inflammatory bowel disease (IBD), or psoriasis and spondyloarthropathies.

But within the group of rheumatoid arthritis patients taking biologics, infliximab was associated with significantly more infections than was etanercept or adalimumab (adjusted hazard ratio 1.25). "This observation may have important implications for the interpretation of studies that report on tumor necrosis factor [TNF]–alpha [antagonists as a group, perhaps because the prevalence of infliximab use could have influenced observed associations," wrote Dr. Carlos G. Grijalva of Vanderbilt University, Nashville.

The drug regimens studied included the use of infliximab-based regimens, many of which used methotrexate to inhibit anti-infliximab antibodies. "Disentangling the effect of individual drugs when used concurrently is difficult. Nevertheless, we noted that concurrent methotrexate use was similar for the other TNF-alpha antagonists."

Also, in a strong dose-dependent fashion, glucocorticoid use at baseline significantly increased the infection risk for patients with rheumatoid arthritis or psoriasis and spondyloarthropathy, Dr. Grijalva and his colleagues wrote in a study posted online in JAMA on Nov. 6 (2011 Nov. 6 [doi:10.1001/jama.2011.1692]).

Dr. Grijalva and his coinvestigators based their findings on data extracted from four large U.S. databases – three included Medicaid and Medicare participants and one included pharmaceutical assistance programs for the elderly and disabled – as part of the Safety Assessment of Biologic Therapy (SABER) project.

Outcomes were assessed within the first 365 days of initial therapy with the TNF-alpha antagonists infliximab, adalimumab, or etanercept and with nonbiologic therapies specific to each disorder.

For rheumatoid arthritis, the comparator therapies were leflunomide, sulfasalazine, and hydroxychloroquine after any use of methotrexate in the previous year. For inflammatory bowel disease, the comparators were azathioprine and mercaptopurine. For psoriasis and the spondyloarthropathies, the comparators were methotrexate, hydroxychloroquine, sulfasalazine, and leflunomide.

The cohort consisted of 10,484 matched pairs of rheumatoid arthritis patients, 2,323 with inflammatory bowel disease, and 3,215 with psoriasis or a spondyloarthropathy. Overall, 20% of patients were aged 65 years or older. Over the 1-year follow-up period, 225 patients died; of these, 148 had rheumatoid arthritis, 38 had inflammatory bowel disease, and 39 had psoriasis.

There were 1,172 serious infections requiring hospitalization, and 53% were for pneumonia or skin and soft tissue infections. The death rate was 3.6% for RA patients, 2% for IBD patients, and 7% for those with psoriasis and the spondyloarthropathies.

Among those with RA, the infection rate was 8.16 per 100 person-years for those taking a biologic agent and 7.78 per 100 person-years for those on a nonbiologic agent, a nonsignificant difference.

However, the authors said, there was a significant difference in infections when considering glucocorticoid use. Those who used up to 5 mg/day were 32% more likely to be hospitalized; 5-10 mg/day, 78% more likely; and more than 10 mg/day, three times more likely to have a hospitalization, compared with patients not taking a glucocorticoid.

When the investigators broke the biologics down into individual infection associations, infliximab was associated with a significant increase, compared with the nonbiologics (adjusted HR 1.26). Infliximab was also associated with significantly more hospitalizations than either etanercept or adalimumab (adjusted HR 1.26 and 1.23, respectively).

Among those with IBD, the rate of infection did not differ significantly between those taking nonbiologic and biologic drugs (9.6 per 100 person-years and 10.9 per 100 person-years, respectively). Nor was there a significant association of infection with baseline glucocorticoid use.

Among those with psoriasis or a spondylarthropy, the infection rates were 5.37 per 100 person-years for nonbiologics and 5.4 per 100 person-years for biologics – not significantly different. But again, baseline glucocorticoid use significantly affected these rates. Those taking 5-10 mg/day of glucocorticoids were twice as likely to be hospitalized as those not taking the drug. Those taking more than 10 mg/day were 2.7 times more likely to have a hospitalization.

Compared with prior studies, this analysis found higher rates of infection overall, a finding that might be due to the large numbers of low-income, elderly, and disabled patients – particularly vulnerable groups.

The researchers said the findings are limited by reliance on coded information from claims and limited information on actual use of medications. The study was primarily sponsored by the Food and Drug Administration and the Department of Health and Human Services. Dr. Grijalva did not have any financial disclosures, but eight of the study investigators reported multiple associations with pharmaceutical companies.

Recommended Reading

Measures of RA Progression Remain Flawed Predictors of Outcomes
MDedge Rheumatology
RA Patients Have Made Treat-to-Target Personal
MDedge Rheumatology
Biologics Can Change the Face of RA
MDedge Rheumatology
RA: Atypical Heart Disease Presentation Can Be Fatal
MDedge Rheumatology
Psoriatic Arthritis: Evidence Lacking for Widespread Methotrexate Use
MDedge Rheumatology
Study: Altered Gait May Flag Early Knee OA
MDedge Rheumatology
Depressed Mood, RA Disease Activity Linked
MDedge Rheumatology
Researchers Confirm PsA Susceptibility Allele
MDedge Rheumatology
New Protein Marker Predicts Success of JIA Treatment
MDedge Rheumatology
Hip Surgery: Revision Risk Higher if Age <75
MDedge Rheumatology