Infections in RA patients treated with infliximab or etanercept Joachim Listing 1 , Anja Strangfeld 1 , Rolf Rau 2 , Ulrich von Hinueber 3 , Maria Stoyanova-Scholz 4 , Erika Gromnica-Ihle 5 , Angela Zink 1 1 German Rheumatism Research Centre, Berlin, 2 Ev. Fachkrankenhaus, Ratingen, 3 Rheumatologist, Hildesheim, 4 Klinik für Rheumatologie, Duisburg, 5 Rheumaklinik Berlin-Buch, Berlin, all Germany Introduction To investigate the long-term safety, effectiveness and costs of biologic therapies in rheumatoid arthritis (RA) the German Society of Rheumatology invited all rheumatologists to contribute to a national prospective cohort study in 2001. The study is known as RABBIT which is a German acronym for: rheumatoid arthritis – observation of biologic therapy. The data were used to estimate the incidence rates of serious and non-serious infections in RA patients treated with etanercept or infliximab and to contrast these rates to those observed in patients treated with conventional DMARDs. Conclusion: Patients treated with biologic agents have a higher a-priori risk of infections. However, our data suggest that this risk is further increased by the treatment with TNF inhibitors. www.biologika-register.de Significant differences were also found for serious infections (etanercept: 6.4 [4.5-9.1], infliximab 6.2 [4.0 – 9.5], control group 2.3 [1.3 – 3.9]). Higher rates in the biologics groups were especially found for lower respiratory tract infections, bacterial skin infections, bone and joint infections. However, the higher risk in the biologics groups could only partly be attributed to the drugs themselves. The different predispositions of the patients had to be taken into account. E p i d e m i o l o g y U n i t DRFZ German Rheumatism Research Centre, Berlin Patients and Methods Patients  RA patients enrolled into the German biologics register RABBIT  new prescription of etanercept or infliximab  new prescription of a DMARD after at least one DMARD failure (control group). Assessments  Treating rheumatologists assessed adverse events (AE) and serious adverse events (SAE) according to the ICH guidelines.  MedDRA v. 7.0 was used to code the adverse events. Statistical analysis  Inclusion of all AE/SAEs experienced within the first 12 months.  AE/SAE rates per 100 observed patient years were calculated.  Propensity score methods were applied to estimate which part of the increase was attributable to differences in patient characteristics.  The following risk factors were included in the propensity (logistic regression) model: age, number of DMARD failures, rheumatoid factor, disease activity score (DAS28), CRP, and disability measured by the Hannover functional questionnaire (FFbH). These factors indicate a higher likelihood of being treated with biologics as well as a higher susceptibility to infections. Results 1,459 patients were enrolled between May 2001 and September 2003. As expected, the patients in the biologics groups had significantly more active disease and more previous DMARD failures (Tab. 1). Values are means if not otherwise specified. The dropout rate was low (11.1%). In total 483, 325, 571 patient years of follow up were available for patients from the etanercept, infliximab, and control group respectively. The infection rates per 100 patient years were for etanercept, infliximab, and control group 22.6 (95% CI: 18.7 – 27.2), 28.3 (23.1 – 34.7), 6.8 (5.0 – 9.4) (p<0.0001) respectively. 63.4 53.9 52.7 FFbH (mean % of full function) 5.4 6.0 6.1 DAS 28 Comorbid conditions (%) 8.2 6.1 9.5 Chronic lung disease 7.9 7.6 8.1 Diabetes 7.7 10.8 10.5 Swollen joint count (0-28) 56.5 53.6 53.7 Age 2.0 3.4 3.6 No. Previous DMARDs 6.0 601 Controls 8.0 346 Infliximab 9.0 512 Etanercept Disease duration (median, years) N Patients enrolled till September 2003 Disclosure: Supported by a joint, unconditional grant from Wyeth Pharma GmbH, essex pharma GmbH, Amgen GmbH and Abbott GmbH & Co. KG. Adjusted for differences in patient case mix by propensity score methods the relative risk of infections in comparison to the controls decreased for serious and non-serious infections by nearly one third (Tab 4). 0 0.3 0 Tuberculosis 0.9 2.2 2.7 other lower RTI 0.5 2.5 1.2 among them: Pneumonia 1.8 11.4 7.0 Respiratory tract infections (RTI) 1.3 3.4 5.8 Other 6.8 28.3 22.6 AE total 5.1 20.6 15.7 Among them: moderate/severe AEs 1.2 4.0 3.7 Bacterial skin and subcut. tissue infect. 0.7 4.0 2.7 Influenza like illness 0.4 1.4 Control 2.1 1.5 Gastrointest. and oral soft tissue infect. 3.4 Infliximab 1.9 Etanercept Herpes viral infections Adverse events Tab. 4: Relative risk (RR) of infections compared to the control group Tab. 1: Patient characteristics 0.4 0.6 0.6 Herpes viral infections 0.4 0 0.6 Sepsis 0.4 0.6 0.8 Other 2.3 6.2 6.4 SAE total 0.4 1.2 1.5 Bacterial skin and subcut. tissue infect. 0.7 3.4 1.9 Lower respiratory tract infections 0.2 Control 0.3 1.0 Bone and joint infections Infliximab Etanercept Serious adverse events Tab. 3: Serious infections / 100 patient years Tab. 2: Infections / 100 patient years 2.1 3.0 Adjusted RRs 2.7 4.1 Unadj. RRs Infliximab 2.2 2.3 Adjusted RRs 2.8 3.3 Unadj. RRs Etanercept 1.8 - 5.1 1.4 - 3.9 Infections total 0.8 - 5.5 0.9 - 5.4 Serious infections 1.3 - 5.9 1.4 - 5.9 Serious infections 95% CI 95% CI 2.8 - 6.0 2.3 - 4.8 Infections total  By this method subgroups of patients with a comparable likelihood of being treated with biologics were identified and Poisson regression was applied to calculate adjusted relative risks based on these subgroups. Acknowledegment The authors wish to thank those rheumatologists who enrolled at least 25 patients each: S Wassenberg and G Herborn, Ratingen; K Babinsky, Halle; A Kapelle, Hoyerswerda; T Klopsch, Neubrandenburg; W Demary, Hildesheim; G-R Burmester, Berlin; K Rockwitz, Goslar; J Kekow, Vogelsang-Gommern; A Bussman, Geilenkirchen; H-E Schroeder, Dresden; E Edelmann, Bad Aibling; E Wilden, Köln; T Karger, Köln; R Dockhorn, Weener; A Graessler, Pirna; B Krummel-Lorenz, Frankfurt/Main; K Krueger, München; W Liman, Hagen; H Soerensen, Berlin; WL Gross, Luebeck/Bad Bramstedt; C Richter, Stuttgart; M Grünke, Erlangen; W Ochs, Bayreuth; S Schewe, München; L Meier, Hofheim; H Tremel, Hamburg; V Petersen, Hamburg; C Kneitz, Würzburg; M Zaenker, Eberswalde; M Bohl-Bühler, Brandenburg; P Herzer, München; A Gause, Elmshorn; R Haux, Berlin; T Grebe, Attendorn; K Alliger, Zwiesel; T Dexel, München; U Lange, Bad Nauheim; D Pick, Grafschaft Holzweiler; K Karberg, Berlin; K Gräfenstein, Treuenbrietzen; C. Stille, Hannover. View publication stats View publication stats