Conference Coverage

Higher infection risk in RA seen with high blood biologic levels

View on the News

Data are not definitive

In this study, the authors use the major British Society for Rheumatology Biologics Register – Rheumatoid Arthritis and examine infections and serious infections across biologics. They define “low/normal” blood levels versus “high” blood levels based on concentration-effect curves. Examining data censored at 1 year versus incidence during 1 year, the results are somewhat inconsistent. With larger numbers available for data censored at 1 year, there is some increased risk using hazard ratios for both all infections and serious infections. With smaller numbers for incident infections during the first year, this hazard ratio does not show an effect.

Dr. Daniel E. Furst, professor of rheumatology at the University of Washington, Seattle, who also is affiliated with the University of California, Los Angeles, and the University of Florence, Italy.

Dr. Daniel E. Furst

These data are interesting and consistent with logic, with increasing concentration being associated with more infection, across biologics. There are a few issues, however. First, the cutpoint dividing low/normal from high appears quite arbitrary. While it is understandable that the authors combined across biologics to achieve adequate numbers of cases, the varying mechanisms of action make this approach somewhat arbitrary as well. Likewise, the issue of low numbers abrogates any significant hazard ratios when looking at incident cases, the most rigorous approach. This study supports previous data but is, unfortunately, neither significantly new data nor definitive data.

Daniel E. Furst, MD, is professor of medicine (emeritus) at the University of California, Los Angeles, an adjunct professor at the University of Washington, Seattle, and research professor at the University of Florence (Italy). He is also practices part-time in Los Angeles and Seattle.



– Higher blood biologic drug levels in the first year of treatment for rheumatoid arthritis independently increased the risk of any infection by about 50% when compared against low or normal levels in a new observational cohort study, providing support for monitoring biologic drug levels to help to predict infection risk.

Data from the British Society for Rheumatology Biologics Register – Rheumatoid Arthritis (BSRBR-RA) that were presented at the British Society for Rheumatology annual conference showed that the adjusted hazard ratio for any infection occurring within the first year among patients with high drug levels was 1.51, with a 95% confidence interval (CI) of 1.14 to 2.01. The adjustments took into account patients’ age, gender, disease activity score, and use of methotrexate.

There are more than 10 biologics now available for use in rheumatoid arthritis but deciding which to use in a particular patient remains very much “a trial and error approach,” first author Meghna Jani, MBChB, said at the conference.

Dr. Meghna Jani, a National Institute for Health Research Academic Clinical Lecturer in Rheumatology at the University of Manchester (England). Sara Freeman/MDedge News

Dr. Meghna Jani

“From a patient perspective, one of the most important concerns continues to be the risk of serious infections and adverse events,” added Dr. Jani, a National Institute for Health Research Academic Clinical Lecturer in Rheumatology at the University of Manchester (England).

The link between biologic agents and infections, including those that could result in hospitalization or other serious consequences, has been well studied in biologics registries. It is known, for example, that the risk of infections with tumor necrosis factor inhibitor treatment seems to be highest during the first 6-12 months of treatment.

According to Dr. Jani, conventional means of determining risk – such as patient age and the presence of comorbid factors – have limited benefit in terms of deciding which patients could be at heightened risk of infections. “Ideally, we need biomarkers in rheumatology that can be implemented in clinical practice and help us predict efficacy and safety, as well as help us use these medications much more cost-effectively,” she said.

Four years ago, a meta-analysis (Lancet. 2015;386:258-65) suggested that the risk of infection may be linked to using higher doses of anti–tumor necrosis factor drugs, which led the BSRBR-RA team to see if elevated levels of these drugs in the serum could be predictive of the infection risk and thus used as a possible biomarker. There was also prior evidence that serum drug concentrations of biologics were associated with long-term treatment response and that a certain level was needed to determine the likely treatment response.

In the current study, Dr. Jani and colleagues used data on 703 patients with rheumatoid arthritis starting biologic therapy who were simultaneously recruited into the BSRBR-RA, which has been running since 2001, and the Biologics in Rheumatoid Arthritis Genetics and Genomics Study Syndicate (BRAGGSS). The BSRBR-RA did not collect biological samples, but in BRAGGSS serological samples were collected at 3-, 6-, and 12-month intervals after the start of a biologic treatment, along with other assessments. This is the first time two national, U.K.-based, rheumatoid arthritis cohorts have been linked in this way, Dr. Jani said.

Serum samples taken from the patients were assessed via enzyme-linked immunoassay to determine levels of the biologic agent used, with high drug levels defined as more than 4 mcg/mL for etanercept (n = 286), tocilizumab (n = 104), and infliximab (n = 14); more than 8 mcg/mL for adalimumab (n = 179), and 25 mcg/mL or more for certolizumab pegol (n = 120).

In the study, about three-quarters of the patients were women. The mean age was 58 years, and disease duration was just under 6 years. Most patients were starting their first biologic.

The crude rate of all infections at 1 year, including recurrent infections, was 464 per 1,000 patient-years in the high biologic drug level group versus 314 per 1,000 patient-years in the low biologic drug level group. When only the first infections were considered, the crude rate of all infections within the first year were a respective 300 and 229 per 1,000 patient-years, with an adjusted hazard ratio of 1.27, Dr. Jani reported.

As expected, lower respiratory tract infections were the most common type of infection, occurring in 34% of patients with high drug levels versus around 10% in the low drug level group. Upper respiratory tract, urinary tract, and skin infections including shingles were seen in a respective 16%, 15%, and 8% in the high drug level group, with rates less than 5% in the low drug level group.

Of note, there were certain types of infections present in the high but not low drug level groups: bacterial peritonitis, neutropenic sepsis, and herpes zoster.

Crude rates for serious infections at 1 year were 76 and 54 per 1,000 patient-years, respectively, for the high and low drug level groups. The crude rates for the first serious infection within the first year were 44 and 29 per 1,000 patient-years. The adjusted hazard ratio for the risk of serious infection at 1 year was 1.26. Serious infections were rare events, Dr. Jani emphasized, so the power was reduced, but “there was a slightly increased risk.”

Aside from the low statistical power to assess the rarer serious infections, another limitation was that drug levels were not measured at the time of the adverse event.

Concluding, Dr. Jani suggested that perhaps monitoring drug levels could be useful in predicting the risk of infection in patients being treated with biologics for rheumatoid arthritis.

Furthermore, “in patients with remission, dose-tapering guided by therapeutic drug monitoring may help lower infection risk and help us balance safety and efficacy.”

Dr. Tore K. Kvien, department of rheumatology, Diakonhjemmet Hospital, Oslo

Dr. Tore K. Kvien

When asked to comment, Tore K. Kvien, MD, PhD, of the department of rheumatology at Diakonhjemmet Hospital in Oslo, supported this conclusion. “Therapeutic drug monitoring [TDM] is widely used among gastroenterologists when treating inflammatory bowel diseases with TNF inhibitors. In recent years, data from several research groups in rheumatology have indicated that TDM may help to optimize drug efficacy. The results from Dr. Jani and her colleagues also support that TDM may be important for safety. The importance of TDM as a ‘new’ hot topic in rheumatology is also supported by the recent establishment of a EULAR [European League Against Rheumatism] task force to further explore the value of TDM when treating patients with inflammatory joint diseases.”

The BSRBR-RA is funded through the BSR, which receives restricted income from several U.K. pharmaceutical companies. These currently include AbbVie, Celltrion, Hospira, Pfizer, UCB, and Roche, and in the past, Swedish Orphan Biovitrum and Merck. The pharmaceutical company funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Dr. Jani has no personal conflicts of interest to disclose.

SOURCE: Jani M et al. Rheumatology, 2019 April;58(Suppl 3):kez105.018.

Next Article: