Background/Purpose

The wide use of biological therapies (BTs) has clearly modified the therapeutic approach in rheumatoid arthritis (RA). One of the most commonly used tools to measure the efficacy of BTs in RA is the DAS28 score. Little is known about how each of the DAS28 components varies over time. Our aim was to graphically evaluate the evolution of the DAS28 components over time for the most common anti-TNF treatments and to compare them with anti-IL6 therapy.

Methods

222 RA patients treated with BTs during the period between Dec’99 and March’13 were included. The data on the components of DAS28 score were collected from baseline and at every 3 months of therapy. First, we visually analyzed the relative improvement of each variable for each anti-TNF (infliximab (INF), etanercept (ETN), adalimumab (ADA)). Second, we grouped the anti-TNFs and compared their combined evolution to tocilizumab (TCZ). In order to obtain a precise visualization of the changes in time of each of the DAS28 components, we used the radar charts. In this type of multivariate data visualization technique, each one of the components is represented as different circle axes and, at each time point, the mean relative improvement (i.e. percentage of reduction from baseline) for each component is connected by a line

Results

A total of 347 BTs were analyzed (INF=100, ETN=126, ADA=79, TCZ=42). No statistically significant differences were found in the DAS28 changes between the 3 anti-TNFs.  However, when comparing the graphical patterns of these therapies we found different patterns of evolution of the DAS28 components over time (Figure 1). Compared to ADA or INF, ETN had a clearly regular expanding pattern of the improvement. In ADA and INF the relative weight of each DAS28 component at each time period was variable, suggesting a different mode of action for monoclonal therapies 

When we compared the radar charts of the combined anti-TNFs against the anti-IL6 treatment (Figure 2), we found a markedly different pattern of action. While the treatment with anti-TNFs showed rapid and greater improvement in the articular component (i.e. SJC and TJC), the treatment with anti-IL6 showed a higher improvement of the ESR (Figure 2). In both treatments, the physician and patient global assessment showed a similar evolution over time

Conclusion

Using a graphical analysis approach we have identified, for the first time, the differential time patterns of the DAS28 score components for the most commonly used anti-TNF therapies. Combining all anti-TNF therapies into a single entity and comparing to a BT targeting IL6, we have also found key differences in the temporal evolution of each of the DAS28 components. The results of this study are useful to understand the differential mechanisms of action of each treatment and could help to explain the differences observed in clinical trials, meta-analyses or observational studies