Endoscopic control of gastrointestinal hemorrhage by local injection of absolute ethanol: a basic assessment of the procedure.

To investigate the hemostatic mechanism of local ethanol injection, ethanol was injected into the gastric mucosa of five adult mongrel dogs and one guinea pig and histologic changes at acute and healing stages were followed. Following a local injection of absolute ethanol, the blood flow in small vessels at the site of injection became instantaneously arrested. Histopathologically, thrombosis of blood vessels in the mucosa and submucosa with edema, predominantly submucosal, was found 10 min after the ethanol injection. There was little or no inflammatory cell infiltration in the injected region. By 4 days after the injection, the base of the formed ulcer became stabilized with a uniform white coating, which, microscopically, was mainly composed of a thick layer of necrotized mucosal tissue. Ethanol showed fixative activity when applied to tissues at concentrations of greater than or equal to 20% although an ethanol concentration of at least 70% was required to accomplish adequate tissue fixation. With 95% or lower concentrations of ethanol, dehydration of tissue was insufficient, and hemostasis due to local vasoconstriction was less conspicuous than with absolute alcohol. Nevertheless, a fixative effect with consequent degeneration and necrosis of cells and secondary thrombosis was evident even at these ethanol concentrations. In patients treated with this hemostatic procedure, exposed blood vessels in the gastric ulcer became necrotized and corroded by fixation with injected absolute ethanol and disappeared from the base of the ulcer within 24 hr as seen in the animals.