P411. Meta-analysis of the efficacy of a second-line anti-TNF in inflammatory bowel disease patients with failure to a previous anti-TNF treatment
J.P. Gisbert, A. Marín, A. McNicholl, M. Chaparro, Hospital Universitario de La Princesa, IP and CIBERehd, Gastroenterology Unit, Madrid, Spain
Approximately one third of the inflammatory bowel disease (IBD) patients receiving anti-TNF treatment do not respond to treatment (primary failure), and another relevant proportion experiments a loss of response (secondary failure) or intolerance to the treatment. The efficacy of a second anti-TNF in these situations has not been evaluated in depth.
Aim: To investigate the efficacy of a second anti-TNF treatment after failure or intolerance to a first anti-TNF drug.
Study selection: Studies evaluating response or remission of a second anti-TNF after the withdrawal of a previous anti-TNF treatment in patients with Crohn's disease (CD) or ulcerative colitis (UC) were meta-analyzed using the inverse variance method. Inclusion criteria: Studies evaluating the response or remission with infliximab (IFX), adalimumab (ADA) or certolizumab (CTZ) as second anti-TNF. Search strategy: Bibliographical search was performed in PubMed up to August 2013. Outcomes: Percentage of response/remission. Subanalyzed according to type of IBD, anti-TNF drug, reason for withdrawal of the first anti-TNF, induction/maintenance dose of the second anti-TNF, and the time of evaluation of the response.
41 studies were included (35 for CD and 6 for UC; 30 switching IFX→ADA, 4 IFX→CTZ, and 1 ADA→IFX). Overall, the second anti-TNF in CD induced remission in 43% (95% CI = 38–48%; I2 = 75%; 27 studies; 2.345 patients) and a response in 65% (95% CI = 57–73%; I2 = 92%; 26 studies; 1.922 patients) of patients. Remission rate was higher when the reason to withdraw the first anti-TNF was intolerance (61%; 95% CI = 40–82%; I2 = 89%; 225 patients) than after secondary (45%; 95% CI = 34–57%; I2 = 79%; 367 patients) or primary failure (30%; 95% CI = 22–37%; I2 = 8%; 175 patients); response rates were respectively 72%, 66% and 60%. Sub-analysis of IFX→ADA switch showed similar results to the overall analysis. The IFX→CTO switch was not sub-analyzed due to low number of studies. The effect of dosage could only be analyzed in the IFX→ADA after secondary failure, showing higher remission (51% vs 44%) and response (81% vs 66%) with the 180/80 scheme than with the 80/40. Overall, UC remission was 17% (95% CI = 2–32%; I2 = 85%; 151 patients) and response was 47% (95% CI = 34–59%; I2 = 64%; 203 patients).
Efficacy of an anti-TNF after failing to a previous anti-TNF treatment is higher in CD than in UC. In CD, the second anti-TNF is more effective when the first one has been stopped due to intolerance (remission of 61%) than due to secondary (45%) or primary (30%) failure. Although all comparisons were heterogeneous and must be interpreted with caution, all analysis consistently showed that efficacy of an anti-TNF switch depends greatly on the cause for switching.