Twenty years of embolization for acute lower gastrointestinal bleeding: a meta-analysis of rebleeding and ischaemia rates.

Abstract

Transarterial embolization (TAE) for acute lower gastrointestinal bleeding (LGIB) can be technically challenging due to the compromise between achieving haemostasis and causing tissue ischaemia. The goal of the present study is to determine its technical success, rebleeding, and post-embolization ischaemia rates through meta-analysis of published literature in the last twenty years. PubMed, Embase, and Cochrane Library databases were queried. Technical success, rebleeding, and ischaemia rates were extracted. Baseline characteristics such as author, publication year, region, study design, embolization material, percentage of superselective embolization were retrieved. Subgroup analysis was performed based on publication time and embolization agent. A total of 66 studies including 2121 patients who underwent embolization for acute LGIB were included. Endoscopic management was attempted in 34.5%. The pooled overall technical success, rebleeding, post-embolization ischaemia rates were 97.0%, 20.7%, and 7.5%, respectively. Studies published after 2010 showed higher technical success rates (97.8% vs 95.2%), lower rebleeding rates (18.6% vs 23.4%), and lower ischaemia rates (7.3% vs 9.7%). Compared to microcoils, NBCA was associated with a lower rebleeding rate (9.3% vs 20.8%) at the expense of a higher post-embolization ischaemia rate (9.7% vs 4.0%). Coagulopathy (P = .034), inotropic use (P = .040), and malignancy (P = .002) were predictors of post-embolization rebleeding. Haemorrhagic shock (P < .001), inotropic use (P = .026), malignancy (P < .001), coagulopathy (P = .002), blood transfusion (P < .001), and enteritis (P = .023) were predictors of mortality. Empiric embolization achieved a similarly durable haemostasis rate compared to targeted embolization (23.6% vs 21.1%) but a higher risk of post-embolization ischaemia (14.3% vs 4.7%). For LGIB, TAE has a favourable technical success rate and low risk of post-embolization ischaemia. Its safety and efficacy profile has increased over the last decade. Compared to microcoils, NBCA seemed to offer a more durable haemostasis rate at the expense of higher ischaemia risk. Due to the heterogeneity of currently available evidence, future prospective and comparative studies are warranted. (1) Acute LGIB embolization demonstrate a high technical success rate with acceptable rate of rebleeding and symptomatic ischaemia rates. Most ischaemic stigmata discovered during routine post-embolization colonoscopy were minor. (2) Although NBCA seemed to offer a more durable haemostasis rate, it was also associated with a higher risk of ischaemia compared to microcoils. (3) Coagulopathy, malignant aetiology, and inotropic use were predictors of rebleeding and mortality. (4) Routine post-embolization endoscopy to assess for ischaemia is not indicated.