Introduction In patients with rheumatoid arthritis (RA) the reasons for failure or loss of response to infliximab (IFX) are controversial, so far both IFX serum levels and the presence of anti-IFX antibodies (HACA) have been associated with these events. In this study we aimed to determine sensitivity and specificity of one available commercial ELISA assay considering cut-off of DAS28 for IFX levels and HACA in RA patients.
Patients and Methods We assessed 60 serum samples from patients diagnosed with RA treated with IFX (1st line of biological treatment) taken prior to infusion. Patients were classified as responders, based on low disease activity (DAS28< 3.2; > 6 months) or non-responders, based on DAS28> 3.2 with > 1 swollen joint and/ or elevated CRP/ ESR. IFX levels and anti-IFX HACA were measured using a commercial ELISA kit (Progenika™) following manufacturer's recommendations. The sample size was previously calculated to ensure a sensitivity (0.8) and specificity (0.6), with a confidence interval of 95%, related to the clinical activity index DAS28. Statistical analysis to establish the appropriate values for the cutoff was performed using median (percentiles 25-75), U test Mann-Whitney and ROC curves (Receiver Operating Characteristic).
Results Median IFX levels in non-responder RA patients (n = 24) was lower compared with responder patients (n = 36). We identified the optimal cutoff with ROC analysis for DAS28<5.1 with IFX levels as <0.046 ug/ml, with a sensitivity and specificity of 100% and 79%, respectively. Furthermore, the median titers of HACA for non responders (p<0.0001). ROC analysis could not be calculated because of the low percentages of positivity for anti-IFX HACA in the selected patient group (n = 8/60, 13.33%).
Conclusion The cut-off values for IFX levels, rather than anti-IFX HACA determination, may be useful in evaluating clinical response measured by DAS28 in RA. However, there is a priority in standardising laboratory techniques (variability inter/ intra-assay and inter/ intra-laboratory) to validate this information and its possible clinical application.