Clinical Effectiveness of First-Line Anti-TNF Therapies and Second-Line Anti-TNF Therapy Post-Vedolizumab Discontinuation in Patients With Ulcerative Colitis or Crohnʼs Disease: 638

INTRODUCTION: Real-world data is needed to understand and compare the clinical effectiveness of second-line (2L) anti-tumour necrosis factors (anti-TNF) treatment (Tx) post first-line (1L; biologic-naive) vedolizumab (VDZ) and 1L anti-TNF use in patients (pts) with ulcerative colitis (UC) or Crohn's disease (CD). METHODS: This was a real-world, multi-country, multi-centre retrospective study in UC and CD pts (≥18 years old) treated with 1L anti-TNF or 2L anti-TNF after discontinuation of 1L VDZ for any reason (received May 2014-March 2018). Pts were from sites in Canada, Greece and the United States. Anti-TNF therapies included adalimumab, infliximab, golimumab and certolizumab pegol. The index date was defined as date of 1L Tx initiation. Clinical effectiveness data were collected from 1L or 2L Tx initiation to earliest of death, chart abstraction date. Cumulative rates of Tx persistence, clinical response and clinical remission were estimated using the Kaplan-Meier method for UC and CD (separately and combined) over 6 months. Clinical response and remission were assessed using pre-defined hierarchical algorithms of standard disease measures reported in the medical records. P-values were generated using the log-rank test. RESULTS: This analysis included 579 anti-TNF patients (1L: 497 [UC: 224; CD: 273]; 2L: 82 [UC: 58; CD: 24]) from 36 sites. Proportions of patients in each cohort were: adalimumab (1L: 41.4%; 2L: 19.5%), infliximab (1L: 52.7%; 2L: 79.3%), golimumab (1L: 4.8%; 2L: 1.2%) and certolizumab pegol (1L: 0.8%; 2L: 0.0%). Mean (SD) age at index date: 1L, 39.6 (15.2); 2L, 49.4 (18.6) years, male: 1L, 49.9%; 2L, 61.0%, median (range: min-max) disease duration: 1L, 2.0 (<0.1–49.0); 2L, 3.7 (0.1–54.0) years. At 6 months, cumulative rates of Tx persistence (1L: 83.9%; 2L: 83.6%), clinical response (1L: 49.5%; 2L: 65.6%) and clinical remission (1L: 29.5%; 2L: 31.4%), were similar between 1L and 2L patients (Table 1). Results were similar when data were stratified by UC and CD (Table 1), albeit sample sizes were small. CONCLUSION: Cumulative rates of Tx persistence, clinical response and clinical remission were comparable between 1L anti-TNF patients and those who switched to a 2L anti-TNF following the discontinuation of 1L VDZ. Results suggest that 1L VDZ may not impact the effectiveness of subsequent anti-TNF Tx in real-world clinical practice. As 2L sample size was limited, these hypothesis-generating data warrant further study.