Embolization for non-variceal upper gastrointestinal tract haemorrhage: a systematic review

To assess the published evidence on the endovascular treatment of non-variceal upper gastrointestinal haemorrhage.

An Ovid Medline search of published literature was performed (1966-2009). Non-English literature, experimental studies, variceal haemorrhage and case series with fewer than five patients were excluded. The search yielded 1888 abstracts. Thirty-five articles were selected for final analysis.

The total number of pooled patients was 927. The technical and clinical success of embolization ranged from 52-100% and 44-100%, respectively. The pooled mean technical/clinical success rate in primary upper gastrointestinal tract haemorrhage (PUGITH) only, trans-papillary haemorrhage (TPH) only, and mixed studies were 84%/67%, 93%/89%, and 93%/64%, respectively. Clinical outcome was adversely affected by multi-organ failure, shock, corticosteroids, transfusion, and coagulopathy. The anatomical source of haemorrhage and procedural variables did not affect the outcome. A successful embolization improved survival by 13.3 times. Retrospective comparison with surgery demonstrated equivalent mortality and clinical success, despite embolization being applied to a more elderly population with a higher prevalence of co-morbidities.

Embolization is effective in this very difficult cohort of patients with outcomes similar to surgery.