Although the liver metabolizes virtually all ethanol reaching the peripheral circulation, recent studies have suggested that an appreciable quantity of ingested ethanol may be oxidized during absorption through the gastric mucosa [1-3]. Of particular importance are claims that certain histamine-2 (H-2) antagonists inhibit this gastric metabolism and hence cause unexpectedly high blood ethanol concentrations in the imbiber [4, 5]. If correct, such findings have obvious medical, legal, and commercial implications. However, in this issue of the Annals, Raufman and coworkers [6] convincingly demonstrate that none of the commercially available H-2 antagonists influenced the blood ethanol response to ingestion of 0.3 g/kg of body weight of ethanol. This article adds fuel to an ongoing controversy regarding if and how H-2 antagonists influence alcohol metabolism.

Appreciation of the controversy concerning the interaction of H-2 antagonists and ethanol requires an understanding of first-pass metabolism. An ingested compound undergoes first-pass metabolism when an appreciable fraction is metabolized before it reaches the peripheral circulation. Such metabolism may occur in the gastrointestinal mucosa or the liver or both and results from the higher concentration of compound to which these two organs are exposed when delivery is via the gut as opposed to intravenous administration. Comparison of peripheral ethanol levels after oral and intravenous administration of low doses of alcohol clearly demonstrate a first-pass effect [1, 2], and, judging from recently published textbooks [7, 8], the scientific world is embracing the concept that this metabolism takes place in the stomach. In this editorial, I first will present data to support the iconoclastic notion that first-pass metabolism of ethanol occurs in the liver, not the gastric mucosa, and then, with this concept in mind, summarize the many studies that have investigated the influence of H-2 antagonists on ethanol metabolism.

Two types of data have been used to show that the gastric mucosa is the site of first-pass metabolism of ethanol. First, bypass of the stomach via either an intraportal infusion of ethanol in rats [3] or by direct intraduodenal instillation in humans [9] eliminated first-pass metabolism observed with oral dosing. Second, the gastric mucosa has been shown to contain alcohol dehydrogenase (ADH) [1, 2], the rate-limiting step in ethanol metabolism; ADH activity correlates with alterations in first-pass metabolism.

Strong evidence can be marshaled, however, to argue that the stomach is not responsible for the first-pass effect. First, up to 50% of a small ethanol dose may undergo first-pass metabolism in rats [1]. Because only about 20% of ingested ethanol is absorbed from the rat stomach [10], the stomach cannot account for metabolism of 50% of the dose. In contrast, the liver has a first-pass exposure to the entire dose of ingested ethanol and thus could carry out efficient metabolism. Second, ADH activity in the gastric mucosa seemingly is insufficient to oxidize the grams of ethanol required for the first-pass effect. For example, Moreno and Pares conclude that the human stomach could metabolize about 1.2 mmol of ethanol per L in 1 hour [11], a negligible fraction of the approximately 500 mmol/L in a 0.3 g/kg dose. Finally, in-vivo assessment of gastric ADH function determined from measurements of the venous-arterial difference across the stomach for ethanol and its metabolites showed negligible gastric oxidation of ethanol [10]. In contrast, the liver, which has about 100 times the ethanol-oxidizing capacity of the stomach [11], could account for the observed first-pass effect (Levitt MD, Levitt DG. Unpublished data).

If, as argued above, first-pass metabolism occurs in the liver rather than in the stomach, how can one account for the disappearance of this metabolism when the stomach was by-passed by direct portal vein or duodenal instillation of ethanol [3, 9]? The explanation is that first-pass metabolism is extremely sensitive to the rate of hepatic delivery of alcohol; the faster the delivery, the more inefficient is the metabolism. Ingestion of a large dose of ethanol or rapid gastric emptying of a smaller dose saturates hepatic removal and eliminates the first-pass effect. In the experiments in which the stomach was by-passed, the rate of hepatic delivery of ethanol was so rapid that little or no first-pass effect would have been expected.

If gastric metabolism of ethanol is negligible, inhibition of gastric ADH by H-2 antagonists is clinically irrelevant. The possibility remains, however, that these agents could influence ethanol metabolism by altering the gastric emptying rate, hence, the hepatic delivery rate, or by interacting with hepatic ADH. This possibility has been investigated in at least 20 studies by comparing a person's blood ethanol response to a given dose of ethanol, with and without treatment with H-2 antagonists. There is general agreement that treatment with these agents has no appreciable effect on blood ethanol concentrations with doses of ethanol that could lead to legally defined inebriation [12-15]. Thus, there are no legal implications to the concurrent ingestion of H-2 blockers and ethanol.

Controversy does exist, however, concerning a possible interaction with small doses of alcohol. As pointed out by Raufman and colleagues [6], interindividual variability in blood alcohol is marked following ingestion of low doses of ethanol. This variability, in part, probably reflects differences in gastric emptying rate that alter the efficient first-pass metabolism that occurs with slow delivery of ethanol to the liver. As a result, reliable conclusions concerning interactions between H-2 antagonists and low doses of ethanol require large, meticulously controlled trials. A highly publicized study involving a small number of persons initially suggested that ranitidine and cimetidine, but not famotidine, modestly enhanced the blood ethanol response to 0.3 g/kg of ethanol [5]. However, the two largest and best controlled studies [6, 16] of this dose of ethanol in nonalcoholic men (allegedly the most susceptible group) showed that no commercially available H-2 antagonist had an appreciable effect on blood alcohol concentrations.

The possibility still remains that there could be an interaction with even smaller doses of ethanol, and several studies [17, 18] have suggested that H-2 antagonists influence the blood alcohol response to ingestion of 0.15 g/kg of ethanol, the equivalent of about one-half bottle of 6% beer. A carefully controlled trial at this dose indicated that ranitidine increased the peak blood concentration from 4.9 mg/dL to 6.5 mg/dL [17]. Although this increase was statistically significant, it should be stressed that these peak levels only represent an increase from 4.9% to 6.5% of the legal inebriation concentration of 100 mg/dL and that these trivial peak values declined rapidly over 30 minutes. In addition, this difference in peak values is minor relative to the differences that would be observed if this dose of ethanol were taken while fasting (with rapid gastric emptying) compared with ingestion with food.

The purist probably will argue that any increase in blood ethanol is bad. However, in light of the enormous problems represented by the abuse of alcohol and the overuse of H-2 antagonists, it seems safe to conclude that the possible harmful interaction of these agents should be relegated to the status of a nonproblem.