Achievement of individual important response measured by Das28dcrit in active rheumatoid arthritis when treated with tocilizumab: Data from a large prospective observational study

Background: Normal fluctuations in results of disease activity measurements due to short-term situational effects and measurement errors are issues for evaluation of individual clinically meaningful therapeutic response in daily practice in patients with rheumatoid arthritis (RA). To address this aspect, we established earlier a statistical approach to determine a critical difference (dcrit) that defines valid criterion for response as assessed by the Disease Activity Score-28 joints (DAS28)1. With this study DCRIT criterion was used to measure clinical meaningful response in RA-patients treated with Tocilizumab (TCZ) in routine care. Methods: The patient population was derived from a prospective noninterventional study in patients with active RA treated with TCZ. A total of 635 patients were analysed. 310 patients with stable disease and stable therapy (no change in TCZ, cDMARD, NSAID and prednisolone) for at least one year (from month 12 to 24 after TCZ initiation) were used to evaluate the previous established DAS28dcrit cut-off of 1.8 for response. To evaluate individual variations in DAS28 scores despite stable mean DAS28 without treatment changes, we subjected DAS28 scores at 12, 18 and 24 months to an ANOVA model to establish a 95% one sided confidence interval (95% CI) for normal fluctuations; this value was used to define the dcrit for individual changes. The total cohort was used to calculate percentage of achieving DAS28 dcrit-response. Results: The mean baseline disease activity (DAS28) from the stable cohort was 5.52 and decreased to 2.51 (month 12). For the next treatment year, the mean DAS of the entire cohort was stable (month 15 2.44, 18 2.40, 21 2.44 and 24 2.46). The calculation of the dcrit cut-off in TCZ-monotherapy results in 1.81 which is within the range of dcrit values in other treatment regimes; for the population including combinational therapy, result was slightly higher (1.9). 70.5% of the patients achieved DAS28dcrit, reflecting meaningful clinical response only arised from the treatment intervention after 12 weeks. The responders increased up to 76.0% at week 24 and 78.9% at week 52 (calculated based on dcrit cut-off 1.8). After 2-years of treatment 80.6% of the patients reach a stable DAS28dcrit response. If using the more conservative (1.94) the result reads as follows: 64.5% week 12, 73.8% week 24, 77.3% week 52, 77.6% at week 104. Conclusions: The established critical differences for DAS28 (DAS28dcrit) of 1.8 was confirmed in an independent cohort of TCZ-treated patients. The slightly higher value of critical difference in TCZ treated patients (especially in TCZ+DMARD) might be due to the unique effect of TCZ on one variable of DAS28 (ESR/CRP). For both cut offs (1.8 and 1.9) high percentages of patients achieved this level of response after initiation of TCZ-treatment