P449 Systematic review with meta-analysis: risk of lymphoma associated with anti-tumour necrosis factor (TNF) agents in inflammatory bowel disease

Abstract

Abstract Background We performed a systematic review and meta-analysis to investigate the risk of lymphoma in patients with IBD treated with anti-TNF agents and thiopurines, alone or combined. Methods We conducted a systematic review of studies including IBD patients treated with anti-TNF and/or thiopurines from Pubmed, Embase and Cochrane databases until May 1, 2019. Two methodology reviewers (AC, CD) separately and repeatedly screened and selected studies according to study selection criteria, collected data and assessed quality. The meta-analysis used the mixed-effect Poisson models (particularly generalised linear mixed models, GLMM) to estimate incidence rate ratios (IRRs) in patients unexposed and exposed to anti-TNF, thiopurines and combotherapy (anti-TNF and thiopurines). Results We identified 4 998 citations. Fourteen studies were eligible for the meta-analysis. There were 8 observational studies involving 265 171 patients and 6 randomised controlled trials involving 2 311 patients (table). As compared with unexposed patients, those exposed to anti-TNF, thiopurines and combotherapy had IRRs (per 1000 patient-years) of 1.66 ([CI 95%, 1.18–2.35], p = 0.004 with I2 = 0%), 3.32 ([CI 95%, 1.87–5.90], p < 0.001 with I2 = 66.83%) and 2.64 ([CI 95%, 1.17–5.93], p = 0.02 with I2 = 1.69%), respectively. The risks of lymphoma did not differ between anti-TNF and thiopurines, between combotherapy and anti-TNF and between combotherapy and thiopurines (table). The results were unchanged when the meta-analysis was restricted to observational studies. Conclusion These results show that the risk of lymphoma is higher in IBD patients treated with anti-TNF monotherapy and combination therapy compared with patients unexposed to anti-TNF or thiopurines.