Background: Disease activity scores such as Clincal Disease Activity Index (CDAI), Simplified Disease Activity Index (SDAI), Disease Activity Score 28 joints with Erythrocytes Sedimentation Rate and 4 variables (DAS28-ESR) or with C-reactive protein and 4 variables (DAS28-CRP) are essential to guide treatment decisions in rheumatoid arthritis (RA) and are widely used in research and clinical practice. Since these instruments rely on different components and thresholds, frequent discrepancies in disease activity classification occur. Moreover, some scores that incorporate acute-phase reactants, may be influenced by treatment that modulate C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) independently of clinical improvement. How these scores perform in response to different RA therapies has not been systematically analyzed and comparative evidence is still lacking.

Objectives: This study aims to quantify how the choice of commonly used disease activity scores differs in capturing transitions between disease activity state (Remission, Low Disease Activity - LDA, Moderate Disease Activity - MDA, High Disease Activity - HDA) over time. Our aim is then to assess whether these transitions are differentially affected by treatment (JAK inhibitors -JAKi, Tumor Necrosis Factor inhibitors -TNFi and Interleukin 6 inhibitors -IL-6).

Methods: Patients from 12 RA registers from the international “JAK-pot” collaboration, starting JAKi, TNFi, or IL-6i were included. Treatment groups were propensity score–matched at the treatment course level at baseline. Analyses were restricted to patients with more than one clinical visit and complete data for all four disease activity measures. Disease activity states and transitions were then compared using multi-state Markov models on three outcomes: the mean time spent in each state, the hazard of transitioning between states, and the one-year transition probabilities.

Results: Among 13’498 patients contributing 78’864 medical visits and 20’448 treatment courses, 7’009 patients (45’928 medical visits, 12’874 treatment courses) were eligible for matching. 3’085 patients on IL-6i were matched to similar patients on JAKi (out of 6’599) and TNFi (out of 10’763). Patients could contribute to more than one treatment course across different treatment groups. Disease activity scores provided the same disease activity state for only 13’678 visits out of 45’928 (29.8 % of all visits). Among the individual scores, CDAI had the most visits with high agreement (i.e., two other scores matched its disease activity state), the fewest visits with no agreement (no other scores agreed with it) and the smallest difference in mean time spent in a particular state (remission, LDA, MDA and HDA) across all three classes of treatments (see figure 1). DAS28-ESR displayed the greatest variation (see table 1). When using DAS28-ESR and DAS-CRP, the probabilities of reaching remission were much higher (around 50%) than when using CDAI and SDAI (around 25%). Furthermore, DAS-ESR displayed more transitions from any disease activity states to remission specifically for IL-6-inhibitors, and to a lesser degree JAK-inhibitors, compared with TNF-inhibitors. Considering treatments, mean time spent in remission, hazard of transitioning in remission and probability of reaching remission at one year were all higher with IL-6-inhibitors when using scores including acute phase reactants (DAS28-ESR, DAS28-CRP, SDAI). For JAK-inhibitors, the effect of using a score with acute phase reactants also tended to classify patients in remission more easily, but to a lesser extent than with IL-6-inhibitors.

Conclusions: To adequately compare treatments, disease activity state should depend on patients’ disease evolution irrespective of the treatment administered. The findings of this study underscore that disease activity scores provide very different evaluation of patient’s states, especially when patients are treated with IL-6i and, to a lesser extent, JAKi. We suggest using a disease activity score that is independent of acute phase reactants, such as the CDAI for comparative effectiveness research and for “treat to target” patient care.

Figure 1.

Table 1.

REFERENCES: NIL.

Acknowledgments: NIL.

Disclosure of Interests: Ophelia Zimmermann: None declared, Emilie Laügt: None declared, Denis Mongin: None declared, Romain Aymon: None declared, Denis Choquette: None declared, Louis Coupal: None declared, Catalin Codreanu: None declared, Florenzo Iannone Abbvie, Alfasigma, Amgen, Astra-Zeneca, Csl-Vifor, GSK, Janssen, Novartis, Lilly, UCB, Abbvie, Amgen, Astra-Zeneca, GSK, Janssen, Lilly, UCB, Roberto F. Caporali AbbVie, Alfasigma, Astra-Zeneca, GSK, Eli Lilly, Galapagos, Janssen, Novartis, Pfizer and UCB, AbbVie, Alfasigma, Astra-Zeneca, GSK, Eli Lilly, Galapagos, Janssen, Novartis, Pfizer and UCB, Tore K. Kvien Grünenthal, Janssen, Sandoz, AbbVie, Gilead, Janssen, Novartis, Pfizer, Sandoz, UCB, AbbVie, BMS, Galapagos, Novartis, Pfizer, UCB, Sella Aarrestad Provan: None declared, Burkhard F. Leeb Sandoz GmbH, Eli Lilly Austria GmbH

Pfizer Corp. Austria GmbH

Abb Vie Austria GmbH

Sandoz GmbH, Ruth Fritsch-Stork: None declared, Galina Lukina: None declared, Dan Nordström Pfizer, UCB, BMS, Lilly, MSD, Novartis, Pfizer, UCB, BMS, MSD, UCB, Nina Trokovic: None declared, Karel Pavelka AbbVie, Eli Lilly, Sandoz, UCB, Medac, Pfizer, Jakub Závada Abbvie, Elli-Lilly, Sandoz, Novartis, Egis, UCB, Sanofi, AstraZeneca, Sobi, Abbvie, Novartis, AstraZeneca, Glaxo, Elsa Vieira-Sousa: None declared, Ziga Rotar Abbvie, Pfizer, Eli Lilly, SOBI, Novartis, Astra Zeneca, Stada., Prodromos Sidiropoulos Abbvie, Pfizer, Lilly, Novartis, UCB, Abbvie, Pfizer, Lilly, Novartis, UCB, MSD, Roche, Amgen, GSK, BOEHRINGER-INGELHEIM, ASTRA ZENECA, Janssen, Sandoz, Biocon, Antonios Bertsias: None declared, Axel Finckh AbbVie, BMS, Eli-Lilly, Gilead, MSD, Pfizer, Alfasigma, Astra Zeneca, UCB, AbbVie, Eli-Lilly, Galapagos, Pfizer, Alfasigma, Delphine S Courvoisier: None declared, Kim Lauper Abb Vie, Pfizer, Novartis, AbbVie, Eli-Lilly, Galapagos, Pfizer, Alfasigma.