Patients

Data records of 553 patients with rheumatoid arthritis, fulfilling the American College of Rheumatology (ACR) 1987 revised criteria,⁹ and who have been successfully treated with infliximab, were retrospectively reviewed to select patients who responded to the treatment and switched to etanercept because of the occurrence of adverse events. Patients were identified as eligible for the study according to the following criteria: (1) a Disease Activity Score computed on 44 joints (DAS-44) of <2.4 or improvement of DAS-44 according to the European League Against Rheumatism (EULAR) response criteria¹⁰ at the time of switching; (2) infliximab should have been stopped after the period of drug induction, not before 14 weeks; and (3) switching to etanercept. The exclusion criterion was a lack of clinical response to infliximab.

Patients were treated with infliximab at the standard dosage of 3 mg/kg intravenously at 0, 2 and 8 weeks, and every 8 weeks thereafter in association with a disease-modifying antirheumatic drug (methotrexate) as prescribed by their own rheumatologist. Etanercept was given at a dose of 25 mg twice weekly by subcutaneous injection without changing the current associated disease-modifying antirheumatic drug.

Study design

This retrospective study was carried out in 12 rheumatology units, in Italy, that participated in A New Traffic Approach Regarding Energy Saving project, a nationwide registry of biological agents endorsed by the National Public Health Service. Ethics committees of all participating centres approved the study after the patients gave written informed consent. Clinical data were gathered during a period of 6 months (24 weeks) of etanercept treatment and for the same period of infliximab treatment before infliximab was stopped. EULAR and ACR response criteria¹¹ were assessed every 8 weeks by evaluating DAS-44, including Ritchie Articular Index (RAI), swollen joint count, erythrocyte sedimentation rate (ESR, mm/1st hour), C reactive protein (CRP, mg/dl) and patient global assessment, and by evaluating tender joint count, physician global assessment, patient Visual Analogue Scale (VAS) of pain and patient Health Assessment Questionnaire (HAQ).

Statistical analysis

Values are shown as medians with minimum and maximum range. For statistical analysis, the Kruskal–Wallis test was used to compare medians. χ² test was carried out to compare the frequency of categorical data. The significance level was set at p<0.05. No inferential analysis was performed for missing data.

Patients

In all, 37 of 553 patients with rheumatoid arthritis were included in the analysis (table 1⇓). The time elapsed after infliximab was stopped and before switching to etanercept ranged between 2 and 4 weeks. Follow-up time points before switching are represented as −8, −16 and −24 weeks, and are matched to the same points after switching. It is noteworthy that the design of this study, whereby patients with rheumatoid arthritis with disease remission have been selected, can itself bias the results, and therefore clinical scores may achieve high rates.

Clinical efficacy

All patients taking infliximab had an improvement of DAS-44 according to the EULAR response criteria at the time of switching, as defined by the inclusion criteria to the study. As fig 1⇓ shows, the percentage of patients with improvement in ACR50 clinical response progressively increased during infliximab treatment and further augmented after switching to etanercept. The proportion of patients reaching the ACR50 clinical response after taking infliximab at “−8 weeks” was 53%, and although that of the same patients taking etanercept at both 16 and 24 weeks was 73%, the difference was significant (p<0.05). Likewise, the ACR70 criteria response showed a similar profile, and the percentage of patients reaching ACR70 response after “etanercept at 24 weeks” (50%) was significantly higher (p<0.05) than that of “infliximab at −8 weeks” (30%).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Percentage of patients reaching American College of Rheumatology (ACR)50 (upper panel) and ACR70 (lower panel) clinical response criteria during infliximab and etanercept treatment. Infliximab “−24, −16 and −8 weeks” refers to the time of retrospective analysis before drug withdrawal and switching. NS, not significant.

Figure 2⇓ summarises median values with 25th and 75th centiles of DAS44, HAQ, RAI, VAS pain, ESR and CRP. Comparisons between the different etanercept time points and the last infliximab point before switching have been considered. DAS44 joints was 2.7 at “infliximab −8 weeks”, 2.3 at “etanercept 8 weeks”, 2 at “etanercept 16 weeks” and 1.8 at “etanercept 24 weeks”, without statistically significant differences. Similarly, the value of HAQ remained substantially unchanged after “infliximab at −8 weeks” (0.75), “etanercept at 8 weeks” (1), “etanercept at 16 weeks” (0.9) and “etanercept at 24 weeks” (0.75). Also, RAI did not show statistically significant differences when etanercept 8, 16 and 24 weeks were compared with “infliximab at −8 weeks” (2.5, 2, 1.5 and 3, respectively). VAS pain was significantly decreased at “etanercept 8 weeks” (29), at “etanercept 16 weeks” (29) and “etanercept 24 weeks” (24) in comparison to “infliximab at −8 weeks” (40, p<0.05). ESR values were comparable during etanercept treatment (19, 15 and 14 mm/h at 8, 16 and 24 weeks, respectively) and before switching at “infliximab −8 weeks” (21 mm/h). An analogous behaviour was detected for CRP, whose median levels ranged between 0.6 mg/dl at “infliximab −8 weeks” and 0.4 mg/dl at “etanercept 24 weeks”.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

 Disease Activity Score measured on 44 joints (DAS44), Health Assessment Questionnaire (HAQ), Ritchie Articular Index (RAI), patient Visual Analogue Scale (VAS) of pain, erythrocyte sedimentation rate (ESR, mm/1st hour) and C reactive protein (CRP, mg/dl) during infliximab and etanercept treatment. Infliximab “−24, −16 and −8 weeks” refers to the time of retrospective analysis before drug withdrawal and switching. Box plots represent the median with 25th and 75th centiles and whiskers 10th and 90th centiles. NS, not significant.