AB0779 Comparison of TNF-α inhibitors efficacy between rheumatic diseases and inflammatory bowel diseases

Abstract

BackgroundThe features of the underlying immune-mediated disease can affect the efficacy of TNF-α inhibitors (TNFi). No previous studies compared head-to-head the efficacy of TNFi between rheumatic diseases (RD) and inflammatory bowel diseases (IBD).ObjectivesTo compare the efficacy of TNFi in the treatment of RD (ankylosing spondylitis (AS) and rheumatoid arthritis (RA)) and IBD (Crohn’s disease (CD) and ulcerative colitis (UC)).Methods123 patients were included into the study. 70 of them had RD (50 (40.7%) patients with AS, 20 (16.3%) with RA) and 53 patients had IBD (38 (30.9%) with CD and 15 (12.2%) with UC). The patients received infliximab (INX, 13 (18.6%) with RD and 23 (43.4%) with IBD), adalimumab (ADM, 36 (51.4%) with RD and 13 (24.5%) with IBD) or certolizumab pegol (CZP, 21 (30%) with RD, 17 (32.1%) with IBD). The response to the treatment was evaluated by ASAS20 for AS, EULAR criteria for RA, decrease in CDAI of ≥100 points for CD and Mayo Score decrease for UC. Statistical analysis was performed using the standard methods of descriptive statistics, Chi-squared test and Fisher’s exact test for categorical variables, incidence rate and incidence rate ratio calculation, Student t test for continuous variables. Survival analysis was performed using Kaplan–Meier method and Cox proportional hazards regression. Log rank test was used to statistically compare the curves and p value was calculated.Results95 (77.2%) patients were still responding to TNFα inhibitor after 23 [8; 24] months of observation. 10 (14.3%) patients in RD group and 18 (33.96%) patients in IBD group had loss of response to the treatment. Thus, the response rate of TNF inhibitor in RD patients was 1.3 times higher than among IBD patients (OR 1.298, 95%CI 1.046 to 1.61, p=0.016). Similarly, patients with AS responded to the treatment 1.3 times more often than patients with CD (86% vs 65.79%, OR=1.307, 95%CI 1.013 to 1.687, p=0.025). The efficacy rate of INX and ADM was comparable between the RD and IBD groups (23.1% vs 34.8%, for INX and 16.7% vs 30.8% for ADM, p>0.05) while loss of response to CZP developed more frequently among patients with IBD that with RD (35.3% vs 4.8%, OR 1.472, 95%CI 1.023 to 2.0, p=0.031).Incidence rate (number of cases of inefficacy per 100 patient-years) of TNFα inhibitor inefficacy was much higher in IBD than in RD (5.43 vs 17.02 cases per 100 patient-years, IRR 3.13, 95%CI 1.37 to 7.59, p=0.0033). The same ratio was observed between CD and AS patients (16.92 vs 5.29 cases per 100 patient-years, IRR 3.19, 95%ДИ 1.19 to 9.46, p=0.012).Median time to loss of response was 28.00±8.08 months in RD patients and 21.11±2.63 months in IBD patients (p>0.05). Only among patients taking INX the duration of therapy at the time of loss of response was significantly shorter in IBD group than in RD group (26.0±4.62 vs 53.67±13.37 months, p=0.03, Student t test 2.575). Event-free survival of the treatment (maintenance of efficacy during the whole period of observation) was significantly higher among RD patients than IBD patients (Log-rank test=8.1, plog-rank=0.0044). Cox proportional hazards regression was used to analyze the time to inefficacy, and a hazard ratio of 2.999 (95%CI 1.35 to 6.61, p=0.0052) was reported for IBD patients. This means that at any random time point patients with IBD have 74.9% probability of developing loss of response ealier than patients with RD.ConclusionTNFi are more efficient for the treatment of AS and RA than for IBD (CD and UC).Disclosure of InterestsNone declared