Poor outcome in rheumatoid arthritis patients switching from etanercept to tocilizumab therapy (#341)
At present, there is no standard by which to pre-qualify rheumatoid arthritis (RA) patients for tocirizumab therapy among those who have failed previous anti-cytokine therapy (non-naïve). In this study, forty-one biologic non-naïve patients who had failed one to three prior treatments with methotrexate combined with anti-TNF-a therapy were enrolled in this study. Prior to receiving tocilizumab treatment, serum cytokine, chemokine and soluble receptor levels were measured using Bio-Plex human cytokine/chemokine 27-Plex panel and Millipore soluble receptors panel (sIL-6R, sgp130, sTNFR-I and sTNFR-II). Patients were treated with 8mg/kg of tocilizumab once every 4 weeks. After 16 weeks of therapy (4 administrations), clinical efficacy was judged based on patients’ disease activity score (DAS28-CRP), and whether patients experienced complete remission or non-remission.
All non-naive RA patients except one showed improvements in their DAS28-CRP score at week 16 of therapy and 9 out of 40 (22.5%) non-naïve patients experienced clinical efficacies that were judged as complete remission. The remaining 31 patients experienced non-remission (symptom levels: low=6, median=21, and high=4). In the same study, complete remission (DAS-28-CRP≧2.3) was achieved by 56 % of naïve patients. Although 20 out of 21 patients who switched form etanacept to tocirizumab had improved DAS28 score only two of these patients achieved remission. On the other hand, 5 of 11 patients switching from infliximab reached remission. Single logistic regression analysis revealed that sgp130 and sIL-6R level played a role in these outcomes.
Additionally in comparing naïve and non-naïve RA patients’ pre-treatment cytokine/chemokine levels, we observed that they were relatively similar except for sTNFR-II, which was higher than in non-naïve patients.
These results suggest that prior RA treatment may influence the outcome of further treatments using different biological drugs. Using prediction biomarkers may help clinicians to select more treatments with more favorable outcomes especially for patients who are switching from one treatment protocol to another.