Background TNF inhibitors are widely prescribed as the first line biologics to date because of a great deal of evidence, however, some patients show more obvious response to IL-6 inhibitor treatment. For the purpose of achieving early remission and efficient medical economics, it would be desirable to be able to predict the efficacy of each anti-cytokine therapy before treatment.
Objectives To establish a scoring method that predicts preferable treatment before starting either IL-6 or TNF inhibitor.
Methods In order to investigate the logical bases, the mRNA levels of IL-6 and TNF-α in the peripheral blood of 45 newly diagnosed RA were analyzed with using expression profiles of mRNA detected by Agilent whole human genome 60K cDNA arrays. 37 patients treated with tocilizumab (TCZ) in our hospital were retrospectively analyzed to elaborate a scoring system for predicting the efficacy. Correlation analysis between the laboratory findings at baseline and the ratio of the DAS28-ESR before and 6 months after treatment (DAS-ratio) was performed. Then, laboratory items that were significantly correlated with the DAS-ratio were selected for parameters constituting a scoring method to predict the efficacy of TCZ treatment. For the initial validation, the baseline data of those 37 patients treated with TCZ and 146 patients treated with TNF inhibitor in our hospital were applied to the scoring method. Finally, we evaluated data of 153 patients treated with TCZ at another two different hospitals to investigate the validity of above scoring method.
Results The expression levels of IL-6 mRNA showed significant negative correlation with TNF-α mRNA in the peripheral blood of RA patients. Correlation analysis between laboratory findings at baseline and DAS-ratio revealed that hemoglobin (Hb), platelet count (Plt), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) were significant in laboratory items. On the basis of these results, a scoring method using these 4 items was proposed to predict TCZ efficacy.
If the baseline data of 37 patients treated with TCZ and 146 patients treated with TNF inhibitors in our hospital were applied to the scoring method, it could differentiate good responders to TCZ (>20 points) from the other, especially, good responders to TNF inhibitor (<20 points). With using the baseline data of 153 patients treated with TCZ at another two different hospitals, a strong correlation (p=0.004) between the score and the DAS-ratio was confirmed.
Conclusions The finding of inverse correlation between mRNA expression of IL-6 and of TNF-α in RA supports a possibility to establish a scoring method to predict dominant cytokine that should be targeted for the treatment. With using our scoring method, it seemed to be possible to predict the dominant cytokine. The usefulness of the scoring system for predicting TCZ treatment was supported by a strong correlation between the score and DAS-ratio with using the data of another two different hospitals.
Disclosure of Interest : None declared