Bevacizumab and Gastrointestinal Perforations

A Review From the FDA Adverse Event Reporting System (FAERS) Database

Thomas A. Wichelmann; Sufyan Abdulmujeeb; Eli D. Ehrenpreis


Aliment Pharmacol Ther. 2021;54(10):1290-1297. 

In This Article


We accessed all available data from the FAERS online, public platform from 2004 to present (July 6th, 2021) utilising the search term "bevacizumab." A total of 74 025 cases of bevacizumab-induced ADR were identified. Cases with the term "perforation" occurring within the GI tract were included in our review. All MedDRA search terms for included perforation locations are shown in Appendix A1. In total, 2874 cases of bevacizumab-induced GI perforations were present in the FAERS database. A 2 × 2 contingency table was developed and the reporting odds ratio of bevacizumab-induced GI perforations was calculated using the total reported cases (see Appendix A3).[6]

Individual cases of bevacizumab-induced GI perforation were reviewed and only cases reported by healthcare professionals were included. Variables available for review on the FAERS platform are shown in Appendix A2. Cases that did not report patient age, gender, location of perforation or indication for bevacizumab use were excluded from the final analysis. All remaining cases were reviewed for duplicates which were removed from the analysis. A duplicate case was identified by using safety report numbers unique to each case. A second step was conducted to identify duplicate cases using the presence of multiple identical identifiers. For example, if two cases had the same patient age, same date of event and same patient weight, then one of these cases was removed from the overall dataset.

Specific information obtained for this review included patient age, patient gender, indication for bevacizumab use and location of perforations. When available, the specific location of perforations were noted. Descriptive statistics for all categorical [N (%)] variables were determined on all patients' demographic, clinical characteristics and outcome (including death). The cases occurring in subjects with the three most common indications for bevacizumab use were individually analysed for location of perforation, average age and clinical outcome (ie mortality). Two-sample unpaired t tests were conducted on age distribution of the three most common indications for use. Z-tests were used to compare the number of patients distributed among the descriptive locations (large intestine, small intestine or gastro-oesophageal) of the three most common indications for bevacizumab use.

Given lack of prior published data, a descriptive review was performed on cases of gastro-oesophageal perforations (reported as "Oesophageal perforation," "Oesophageal ulcer perforation," "Gastric perforation," or "Gastric ulcer perforation") utilising the same variables.

The STROBE checklist was used for the writing of this report.[7]