Background/Purpose: Serum 14-3-3η is a mechanistic marker that is involved in the pathogenesis of RA and is a potent up-regulator of IL-6. We previously reported that baseline 14-3-3η levels were significantly lower in a cohort of 49 established RA patients who achieved better clinical outcomes when treated with Tocilizumab. Recently there has been increased interest in assessing therapy response using alternative measures to DAS that are independent of CRP, including patient reported outcomes such as HAQ-DI. The aim of this study was to validate whether lower baseline levels of 14-3-3η is associated with and independently predicts clinical response in RA patients treated with tocilizumab.

Methods: Serum 14-3-3η levels were measured in a cohort of 106 Japanese patients prior to the initiation of tocilizumab therapy (BL). 14-3-3η positivity was defined by the diagnostic cut-off of ≥ 0.19 ng/ml, and 2 and 4 times that, at 0.40 and 0.80 ng/ml, respectively. Patients were sub-grouped according to changes in HAQ-DI based on the minimal clinically important difference (MCID) of ≥ 0.22, ≥ 0.31, and ≥ 0.49. Group differences were assessed by Mann-Whitney U-test. ROC curve analysis was performed to identify the optimal 14-3-3η cut-point for a ΔHAQ-DI ≥ 0.49. Uni- and multi-variable analysis controlling for baseline HAQ-DI was used to assess 14-3-3η’s independence of other clinical/serological variables at informing patient reported outcomes.

Results: At baseline, 75 (71%) of the 106 patients were 14-3-3η positive. At year 1, 70 (66%), 58 (55%) and 52 (40%) of the 106 patients achieved changes in HAQ-DI of ≥ 0.22, ≥ 0.31, and ≥ 0.49, respectively. Baseline HAQ-DI values were significantly higher in those patients that achieved a Δ HAQ-DI ≥ 0.22 when compared with those that did not; p = 0.009. The HAQ-DI values were also significantly higher at the two other HAQ-DI cut-points (≥ 0.31, and ≥ 0.49), p < 0.0001. As shown in the Figure, serum 14-3-3η levels were significantly lower in patients that achieved HAQ-DI MCID of > 0.31 and >0 .49. In patients who achieved a ΔHAQ-DI ≥ 0.49, ROC analysis returned the best cut-off of ≤ 0.18 ng/ml yielding a specificity of 83.3% (95%CI: 70.7-92.1) and sensitivity of 42.3% (95%CI: 28.7-56.8), LR = 2.5. By univariate analysis, all three 14-3-3η cut-points were significantly associated with better clinical outcomes. Multi-variable modeling controlling for BL HAQ-DI indicated that lower levels of 14-3-3η, age and disease duration, but not CRP, were independent predictors of achieving the HAQ-DI MCID outcomes.    

Conclusion: Lower serum 14-3-3η levels prior to the initiation of therapy independently predicts better patient reported outcomes in patients treated with tocilizumab. The predictive capacity of 14-3-3η at informing therapy response to anti-IL-6 therapy should be investigated across all anti-IL-6 compounds.