King's College London

Research portal

Non-steroidal anti-inflammatory agents and anastomotic leak rates across colorectal cancer operations and anastomotic sites: A systematic review and meta-analysis of anastomosis specific leak rate and confounding factors

Research output: Contribution to journalArticlepeer-review

S. L. Kastora, L. L. Osborne, R. Jardine, G. Kounidas, B. Carter, P. K. Myint

Original languageEnglish
Pages (from-to)2841-2848
Number of pages8
JournalEuropean Journal of Surgical Oncology
Issue number11
Early online date26 Nov 2021
Accepted/In press2021
E-pub ahead of print26 Nov 2021
PublishedNov 2021

Bibliographical note

Publisher Copyright: © 2021 Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

King's Authors


Background: Surgical intervention presents a fundamental therapeutic choice in the management of colorectal malignancies. Complications, the most serious one being anastomotic leak (AL), still have detrimental effects upon patients’ morbidity and mortality. We aimed to assess whether NSAIDs, and their sub-categories, increase AL in colonic anastomoses and to identify whether this affects specific anastomotic sites. Materials and methods: A systematic search of MEDLINE, Cochrane Library,, Web of Science, Science Direct, Google Scholar was conducted between January 1, 1999 till the October 30, 2020. Cohort studies and randomized control trials examining AL events in NSAID-exposed, colorectal cancer patients were included. NSAIDs were grouped according to the 2019 NICE guidelines in non-specific (NS-NSAIDs) and specific COX-2 inhibitors. The primary outcome was AL events in NSAID-exposed patients undergoing operations with either ileocolic, colocolic or colorectal anastomoses. Secondary outcomes included NSAID category-specific AL events and demographic confounding factors increasing AL risk in this patient population. Results: Fifteen studies involving 25,395 patients were included in the systematic review and meta-analysis. Of all anastomoses, colocolic anastomoses were found to be statistically more prone to AL events in the NS-NSAID-exposed population [OR 3.24 (95% CI 0.98–10.72), p = 0.054]. Male gender was an independent confounder increasing AL rate regardless of NSAID exposure. Conclusion: The association between NSAID exposure and AL in oncology patients remains undetermined. Whilst in present work, colocolic anastomoses appear to be more sensitive to AL events, the observed association may be anastomotic site and NSAID-category dependent.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454