Dr. Boulos Haraoui

Learning Objectives: Evaluate the risk of serious infections in RA patients treated with anti-TNF agents Compare the relative risk of serious infections among the 3 commercially available anti-TNF agents Stress the importance of patient education regarding the risk of serious infections when anti-TNF therapy is initiated. This paper reviews the incidence of serious infections in patients treated with anti-TNF agents and included in the BSRBR. The control group is made of patients with RA treated with non-biologic DMARDs and never exposed to anti-TNF agents. These 2 groups differed completely for all baseline characteristics.  The definition of “serious” infections is a valid one and includes death, hospitalisation and the use of IV antibiotics.

The overall incidence rate ratio for the anti-TNF compared to the DMARD group almost reached statistical significance at 1.28 (95% CI: 0.94 – 1.76). When adjusted for age and sex the rate becomes significant at 1.47 (95% CI: 1.07-2.01)

After further adjustments for other important confounding factors such as disease severity, extra-articular manifestations, steroid use, co-morbidity… statistical significance is lost. This analysis led the authors to conclude: “in patients with active RA, anti-TNF therapy was not associated with increased risk of overall serious infection compared with DMARD treatment after adjustment for baseline risk”.

Unfortunately this analysis and conclusion will lead to a sense of false reassurance, since the patients treated with anti-TNF agents are in the high risk group for serious infections compared to the patients with milder disease treated only with DMARDs. These methodologically driven adjustment analyses are not relevant to the practising rheumatologist who has to care for sick patients with multiple risk factors, one of them being the addition of anti-TNF therapy.

Other registries and data bases have shown an increased risk of infections in the anti-TNF treated patients and it has been the recommendations of all prescription guidelines, including the CRA, to educate rheumatologists and patients to the risk of infections and to seek immediate medical attention whenever such situation arises.

Moreover, when they analysed the incidence of opportunistic infections and infections with intra-cellular organisms, all 19 cases were in the anti-TNF group. This should have alerted the authors to a more conservative conclusion.

One interesting observation though, is that all 3 anti-TNF agents carry the same risk for serious infections with however a higher incidence for TB with infliximab.

In conclusion, and contrary to the authors’ conclusion this paper stresses the important risk of serious infections in the RA patients treated with anti-TNF agents.

British Society for Rheumatology Biologics Register Control Centre Consortium, Manchester, UK.

OBJECTIVE: To determine whether the rate of serious infection is higher in anti-tumor necrosis factor (anti-TNF)-treated rheumatoid arthritis (RA) patients compared with RA patients treated with traditional disease-modifying antirheumatic drugs (DMARDs).

METHODS: This was a national prospective observational study of 7,664 anti-TNF-treated and 1,354 DMARD-treated patients with severe RA from the British Society for Rheumatology Biologics Register. All serious infections, stratified by site and organism, were included in the analysis.

RESULTS: Between December 2001 and September 2005, there were 525 serious infections in the anti-TNF-treated cohort and 56 in the comparison cohort (9,868 and 1,352 person-years of followup, respectively). The incidence rate ratio (IRR), adjusted for baseline risk, for the anti-TNF-treated cohort compared with the comparison cohort was 1.03 (95% confidence interval 0.68-1.57). However, the frequency of serious skin and soft tissue infections was increased in anti-TNF-treated patients, with an adjusted IRR of 4.28 (95% confidence interval 1.06-17.17). There was no difference in infection risk between the 3 main anti-TNF drugs. Nineteen serious bacterial intracellular infections occurred, exclusively in patients in the anti-TNF-treated cohort.

CONCLUSION: In patients with active RA, anti-TNF therapy was not associated with increased risk of overall serious infection compared with DMARD treatment, after adjustment for baseline risk. In contrast, the rate of serious skin and soft tissue infections was increased, suggesting an important physiologic role of TNF in host defense in the skin and soft tissues beyond that in other tissues.

PMID: 16868999 [PubMed - in process]