Endoscopic Band Ligation Could Decrease Recurrent Bleeding in Mallory–Weiss Syndrome as Compared to Haemostasis by Hemoclips Plus Epinephrine

Endoscopic Band Ligation Versus Hemoclips Plus Epinephrine

S. Lecleire, M. Antonietti; I. Iwanicki-Caron; A. Duclos; S. Ramirez; E. Ben-Soussan; S. Hervé; P. Ducrotté


Aliment Pharmacol Ther. 2009;30(4):399-405. 

In This Article



Among the 56/218 patients presenting with an UGIB related to a MWS during the study period, 29 were endoscopically treated for primary haemostasis by band ligation (Banding group) and 27 by a combination of hemoclips placement and epinephrine injection (H&E group). The 162 remaining patients were not endoscopically treated because of lack of active bleeding or visible vessel on the MW tear. Clinical, biological and endoscopic data at admission are listed in Table 1. There was no significant difference between the two groups with respect to age, gender, shock, haemoglobin level, coagulopathy and the use of antiplatelets or anticoagulant drugs.


Spurting or oozing vessel was present at endoscopy in 55% of patients in Banding group and 56% in H&E group (N.S.) (Figure 1). Primary haemostasis was obtained in all patients in each group (Figure 2). In patients from Banding group, the mean number of elastic bands applied was 1.14, with only five patients requiring the placement of a second band in the same endoscopic session. In patients from H&E group, the mean number of hemoclips applied was 2.3, ranging from one to five hemoclips and the mean volume of injected epinephrine solution was 10 mL, ranging from 4 to 24 mL.

Figure 1.

Bleeding Mallory–Weiss tear.

Figure 2.

Haemostasis by band ligation.


Clinical outcome data are summarized in Table 2. Recurrent bleeding was significantly higher in patients treated by hemoclips placement and epinephrine injection than in patients treated by band ligation (18% vs. 0%; P = 0.02). Rebleeding was successfully treated by endoscopic band ligation in four patients with no further recurrent bleeding. Only one patient did not respond to band ligation as a second haemostasis procedure and underwent a salvage transcatheter arterial embolization, which permitted obtaining a permanent haemostasis. In each case, for patients who experienced rebleeding, an hemoclip detachment was observed. No death was observed in the two groups.

Univariate and Multivariate Analyses of Rebleeding Risk Factors

Univariate and multivariate analyses of rebleeding risk factors were performed using the following variables: the type of endoscopic treatment, an active bleeding at endoscopy, a shock at admission, an antiplatelets or anticoagulants use, an haemoglobin level of less than 10 g/dL at admission. In univariate analysis, rebleeding was significantly associated with an endoscopic haemostasis by hemoclips combined with epinephrine, an active bleeding MW tear at endoscopy, a shock at admission and the use of antiplatelets or anticoagulant drugs (Table 3). The multivariate analysis revealed that risk factors of rebleeding were an haemostasis by hemoclips combined with epinephrine and an active bleeding lesion at endoscopy (Table 3).