Background/Purpose: The biologics therapies in rheumatoid arthritis (RA) implies a well-known risk of infection. It is of utterly importance to count with predictive models to define risk factors for infection. Objectives: Validate the RABBIT risk score for serious infections in a cohort of RA patients followed for 12 months. Evaluate which variables of our population could be of relevance to optimize the score.

Methods: Patients ≥18 years old fulfilling 2010 ACR/EULAR criteria for RA were included and followed for 12 months. Basal sociodemographic data, comorbidities, RA characteristics and vaccination was consigned . Basal RABBIT score was calculed. Serious infections were consigned, describing site and time since enrollment. Statistical analysis: Categorical variables were compared using x2 test and continual variables using Student´s t test or Mann Whitney. Cumulative survival of infection was assessed using Kaplan Meier curves. Variables associated serious infections were evaluated using Cox proportional regression model.

Results: 605 patients were included. 83.5% female, mean age 52.8±12.8 years, 96.2% urban residence, 83% GRAFFAR III/IV (middle, middle-low class), median years of formal studies 10 (IQR 7-13) years. 70% had health insurance. Median time of RA evolution 84 (36-168) months, 90% RF+, 80% ACPA+, median DAS-28 3.5 (2.4-4.8), median CDAI 10 (3-19) and median HAQ 0.87 (0.37-1.50). Comorbidities: 7% Previous infection, 10% lung disease, 1% chronic kidney disease, 6% diabetes. Vaccination status: 67% antiinfluenza, 66.5% antipneumococcal, 47% antiHBV. Treatment: 75% patients were on MTX. 60% were on csDMARDs alone and 40% on bDMARDs (60% TNFi, 20% Tocilizumab, 13% Abatacept, 7% Tofacitinib). 15.6% had a prior history of bDMARDs use. Mean dose of prednisone was 3.6 mg/day and 75% of patients received under 7.5 mg/day. Incidence of serious infection was 5% (CI95% 3-7). Most frequent sites were respiratory and urinary (90%). Mean time to serious infection was 3.5 mont. Variables associated with serious infection: High disease activity by CDAI 4.1 (1.8-9.5) and DAS-28 3.2 (1.3-7.3), previous serious infection 19.3 (8-47), pulmonary disease 5.2 (2.1-13), diabetes 6.4 (2.3-17.6), prednisone >20mg/day 14.3 (4.8-42.5), Tocilizumab treatment 12.4 (3-53) and current use of DMARDs 0.24 (0.1-0.6). Cox regresión model  (serious infection): Previous serious infection (expB:10.6 p:0.00001), diabetes  (expB:4.8 p:0.02), Tocilizumab treatment (expB: 11.4 p:0.012), high dose of corticoids (expB: 9.3 p:0.008) and current use of DMARDs (expB: 0.3 p:0.04) were independently associated. Performance of RABBIT score: Patients with no infection during follow up had a median score of 1.2 (IQR 0.8-2.1) and patients with infection a median score of 5.1 (IQR 2.15-12.6)(p=0.00001). ROC curve analysis: AUC 0.86 (CI95% 0.8-0.94), best cut-off 2.85 (sensibility 75%,  specificity 85%).

Conclusion: The RABBIT score performed excellently in our patients. Serious infections was 5% during the 12 months, mean time to serious infection was 3.5 months and it was independently associated with previous serious infection, diabetes, use of Tocilizumab, prednisone dose >20 mg/day and negatively with current use of DMARDs.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/performance-of-the-rabbit-infection-score-in-a-prospective-multicenter-cohort-of-rheumatoid-arthritis-patients-from-argentina/