Endoscopic treatment and follow-up of gastrointestinal Dieulafoy's lesions.

Aim: To investigate retrospectively the clinical and endoscopic features of bleeding Dieulafoy's lesions and to assess the short- and long-term effectiveness of endoscopic treatment.

Methods: Twenty-three patients who had gastrointestinal bleeding from Dieulafoy's lesions underwent endoscopic therapy. Demographic data, mode of presentation, risk factors for gastrointestinal bleeding, blood transfusion requirements, endoscopic findings, details of endoscopic therapy, recurrence of bleeding and mortality rates were collected and analyzed retrospectively.

Results: Hemostasis was attempted by dextrose 50% plus epinephrine in 10 patients, hemoclipping in 8 patients, heater probe in 2 patients and ethanolamine oleate in 2 patients. Comorbid conditions were present in 17 patients (74%). Overall permanent hemostasis was achieved in 18 patients (78%). Initial hemostasis was successful with no recurrent bleeding in patients treated with hemoclipping, heater probe or ethanolamine injection. In the group of patients who received dextrose 50% plus epinephrine injection treatment, four (40%) had recurrent bleeding and one (10%) had unsuccessful initial hemostasis. Of the four patients who had rebleeding, three had unsuccessful hemostasis with similar treatment. Surgical treatment was required in five patients (22%) owing to uncontrolled bleeding, recurrent bleeding with unsuccessful retreatment and inability to approach the lesion. One patient (4.3%) died of sepsis after operation during hospitalization. There were no side-effects related to endoscopic therapy. None of the patients in whom permanent hemostasis was achieved presented with rebleeding from Dieulafoy's lesion over a mean long-term follow-up of 29.8 mo.

Conclusion: Bleeding from Dieulafoy's lesions can be managed successfully by endoscopic methods, which should be regarded as the first choice. Endoscopic hemoclipping therapy is recommended for bleeding Dieulafoy's lesions.