Experimental studies of injection therapy for ulcer haemorrhage in rabbits

Endoscopic injection therapy is a well established method of controlling peptic ulcer haemorrhage but the optimum injection solution and the mechanism involved in inducing haemostasis are unknown.

The efficacy and effects on tissue of various therapeutic agents used in the control of gastric mucosal haemorrhage were studied in ten rabbits. Thirty-eight bleeding mucosal ulcers (blood loss above 1.5 ml/min) created at gastrotomy were studied. Adrenaline (1:100,000), thrombin, fibrin (thrombin plus fibrinogen), 5 per cent ethanolamine and 50 per cent dextrose were injected; a fibrin suspension was also sprayed around bleeding ulcers.

Sclerosants were found to be least effective in the control of bleeding and were associated with significant tissue necrosis. Although all the other solutions significantly decreased blood loss within 30 min of injection (median blood loss 0.25 ml/min), only an injected mixture of adrenaline plus thrombin and sprayed fibrin achieved complete haemostasis within 2 min of treatment and with no recurrence of bleeding. Neither agent caused significant tissue damage. Histological examination showed that no solution caused arterial thrombosis when injected next to a major ear artery.

Sclerosants caused extensive tissue necrosis and were least effective in the control of ulcer haemorrhage. A combination of dilute adrenaline and human thrombin may represent optimal haemostatic therapy for peptic ulcer haemorrhage.