Drugs that influence gastric fluid pH and volume

This chapter is relevant to Section O2(ii) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "describe the pharmacology of drugs that influence gastric fluid pH and volume". It has attracted a lot of attention from CICM primary examiners, as the practice of offering stress ulcer prophylaxis is ubiquitous in the ICU. It therefore seems logical to impress the importance of these drugs on the juniormost trainees, so that they should grow up cultured. From this reasoning, several questions have sprouted:

• Question 8 from the second paper of 2016
• Question 22 from the first paper of 2011
• Question 10 from the first paper of 2010
• Question 6(p.2) from the first paper of 2008

These are all questions about gastric pH, mainly requiring a comparison of H2 histamine receptor blockers and proton pump inhibitors. Question 8 is even more transparently a naked comparison of ranitidine and omeprazole. In others, examiners made approving noises at trainees who included antacids prostaglandin analogues and anticholinergic drugs.  Each class of this diverse group of substances probably calls for a detailed monograph, but that sort of write-up would strain the patience of even the most stoic reader. Instead of testing that relationship again, a very exam-focused discussion is offered here.

In summary:

Of all the published works that might be suitable for this topic, there is no single resource that's got that magical combination of characteristics (comprehensive, detailed and free). Aihara et al (2003) is probably as good as things get, and has some information about some of the weirder rarer drugs proposed for gastric acid control (eg. cholecystokinin receptor antagonists). 

Classes of drugs used for ulcer prophylaxis

The best way to describe this would probably be in some kind of huge table, listing these substances roughly in order of their availability to the random consumer. This works mainly because they mostly seem to be grouped in classes, as one class became available and dominant over the others. Antacids and atropine gave way to H2 antagonists, which in turn gave way to PPIs. 

Drug/class	Year of availability
Antacids
Calcium carbonate	Antiquity
Bismuth	1865
Sodium citrate	1910 or thereabout
Magnesium oxide	1922
Aluminium hydroxide	1922
Anticholinergic agents
Atropine	1937
Prostaglandin analogues
Misoprostol	1973
Histamine (H2) receptor antagonists
Cimetidine	1971
Ranitidine	1981
Famotidine	1985
Nizatidine	1988
Mucosal protectants
Sucralfate	1981
Cholecystokinin-2 receptor antagonists
Proglumide	1981
Proton pump inhibitors
Omeprazole	1989
Lansoprazole	1992
Pantoprazole	1994
Rabeprazole	1997
Esomeprazole	2000

Briefly, on their chemistry, which is by far the least important aspect of the discussion:

• Antacids are mainly the hydroxides and carbonates of alkali metals
• Anticholinergic agents used for their acid suppressant effects are mainly antimuscarinic alkaloids, like atropine and hyoscine (which are tropane alkaloids and tertiary amines)
• Histamine receptor antagonists are mainly guanidines (Ganellin, 1978). Their pharmacokinetic properties are all very similar because of this.
• Proton pump inhibitors are all benzimidazole derivatives, chemically similar to antihelminthics like albendazole and mebendazole. Of that class, the vast majority of members are encountered mainly in veterinary medicine.
• Sucralfate is a thick weird glorp representing the unnatural union of sucrose sulfate and aluminium hydroxide. If one wishes to be mistaken for a professional, one would refer to it as a basic aluminium salt of sulfated sucrose, which decomposes into a polyanionic gel in acidic environments.
• Prostaglandin analogues such as misoprostol are synthetic analogues of endogenous hormones, and in fact misoprostol was the first such analogue to be made available.
• Cholecystokinin-2 receptor antagonists are actually a rather diverse group of drugs, as people seem to keep developing them over decades, and they never seem to become popular. Proglumide from the 1980s was a glutamic acid derivative for example, and netazepide is a benzodiazepine of all things. There's also quinazolinones and ureido-acetamides and indoles. McDonald (2001) is by far the best effort in explaining this group of substances, most of which were only ever known by their pharma company coded number identifiers. For this reason, they will not be mentioned here. 

Pharmacokinetics of drugs which influence gastric pH

Administration

Antacids are all administered orally, as their mechanism of action involves them mixing with gastric acid directly, which means the rectal route would be too inefficient.

Atropine, when administered for the purpose of raising gastric pH, was usually administered orally; and the vintage of this practice can be discerned from publications like Gill & Jessup (1950), who summarise articles reporting on the effects of "5 or 10 minims of tincture of belladonna". Parenteral administration is absolutely plausible from a pharmacological perspective, in the sense that this drug will have its antacid effect whether administered IV or subcutaneously, but this practice never became popular, perhaps because they were giving doses in the order of 1 mg.

H2 receptor blockers are more marketable as tablet medications, as the number of people who complain vocally of GORD are many, and the intubated GI bleeders are few. As such, companies have focused mainly on oral formulations. In spite of this, intravenous options exist for ranitidine and famotidine.

Proton pump inhibitors are all available as IV formulations as well as oral tablets, though the availability of the IV formulations may vary around the world (for example, IV rabeprazole does not seem to be available in Australia).

Sucralfate is obviously only going to be effective when given orally, as its mechanism of action requires for it to coat eroded gastric surfaces. Realistically, however, it will cake onto all kinds of moist surfaces, and has been used rectally (for radiation proctitis), topically to the genital lesions of inflammatory dermatoses, and on to the surface of skin ulcers and other wounds. 

Misoprostol is available in a variety of routes, including oral and sublingual (not to mention vaginal, for the termination of pregnancy). 

Absorption

Antacids do not need to be absorbed, as they do their job directly at the gastric lumen. However, that does not mean we can completely forget about the A of their ADME.  Firstly, some completely undesirable absorption of their ingredients does take place, which leads to some of their unpleasant side effects (aluminium toxicity, hypernatremia, etc). Secondly, they interfere with the absorption of other drugs, often by binding them directly in some destructive covalent way. Thirdly, even if they are not particularly well-absorbed, they will produce osmotic diarrhoea.

Anticholinergic drugs, of which atropine is a good representative, are usually well absorbed generally, and apparently up to 90% of oral atropine is bioavailable after oral administration (Beermann et al, 1971). 

H2 receptor antagonists have reasonably good bioavailability. For most of them, intestinal absorption is good, and hepatic metabolism is not much of a threat. 

Proton pump inhibitors are well absorbed in the small intestine. Ironically, most of them are rendered useless by gastric acid, which means the oral formulations need enteric coating. The bioavailability is generally quite good (50-80%), and it improves with repeat dosing because of hepatic enzyme inhibition (Fock et al, 2008)

Sucralfate is not absorbed in any meaningful sense, as it is a polymer-like gel which does not tend to break down into easily absorbed molecular chunks. Sure, some aluminium ions or sulfated sucrose disaccharides may break off from the main blob and get accidentally mistaken for something edible by the enterocytes, but this uptake would be negligible and of trivial importance.

Misoprostol is rapidly absorbed and has reasonable bioavailability (~54%); or rather, the effect is not much diminished, because first-pass metabolism produces an active metabolite.

Distribution

Antacids, if you decide to discuss them honestly, are basically elemental inorganic ions with some kind of hydroxide or carbonate group which makes them alkaline. The inorganic ions therefore distribute wherever that ion normally distributes in the body fluids, eg. the sodium in oral sodium bicarbonate gets distributed into the extracellular fluid like any sodium ordinarily would. However, it feels disingenuous to offer this sort of explanation for the distribution of the sodium bicarbonate antacid, because the antacid itself is the sodium bicarbonate molecule rather than sodium. The antacid effect is isolated to the gastric lumen, and the sodium bicarbonate which does all the acid-neutralising work mostly finishes its life there, turning into belch-inducing CO₂ and sodium chloride.

However, antacids (and specifically the sodium bicarbonate in our example) can definitely have a systemic alkalinising effect. In fact oral sodium bicarbonate loading is a known strategy used by exercise enthusiasts to increase their insurance during intense training. In an experiment that administered 300mg/kg (about 250 mmol) of oral sodium bicarbonate to athletes, Hilton et al (2019) observed the standard base excess increase from 0 to 8, and the bicarbonate from 24 to 30 mmol/L.  

There are several ways to explain this. Classical textbooks from respected nonmedical disciplines confidently state that "the unreacted fraction is readily absorbed into the general circulation and may alter systemic pH". By this model, whatever sodium bicarb is left over from its burp-inducing reaction in the stomach will end up getting dumped into the lumen of the intestine, where it can be reabsorbed by the jejunum. From this, it should follow that the same amount of sodium bicarbonate administered in an enteric-coated form should have a greater systemic alkalinising effect, as it does not get wasted by the stomach acid. This, in fact, is not observed. The same study by Hilton et al actually found that an aqueous bicarbonate solution produced an equivalent amount of alkalinisation, but just a bit faster.

This may mean that the bicarbonate itself is not the source of the alkalinising effect. Using a physicochemical model for the analysis of this phenomenon, we could observe that both the rapid aqueous and the delayed enteric-coated formulation would deliver the same amount of sodium to be absorbed by the intestine, and it is the extra sodium  (i.e. the increase in the strong ion difference) that leads to the alkalinising effect.

Atropine, the anticholinergic drug we have the most experience with, has a volume of distribution of about 1-6L/kg, and is 50% protein bound. It has an interesting tissue distribution, rapidly becoming concentrated in various secretory tissues, such as "salivary glands and certain endocrine glands such as the pituitary, thyroid, parathyroid, pancreatic islets and adrenal medulla" (Albanus et al, 1968). It is also found in considerable concentrations in the iris, ciliary body, and the lung, where it persists for some considerable time, whereas it is eliminated rapidly by metabolism from elsewhere.

H2 receptor blockers all have relatively good water solubility and are distributed mainly into body water, with small-ish volumes of distribution (around the 0.8-1.5L/kg). Most of them are 15-30% protein bound, which has minimal clinical relevance. Theoretically all of them should be easily available as IV formulations because of this, but practically their oral administration is so common and convenient that it is often difficult to obtain an IV option.

Proton pump inhibitors vary in their water solubility, from good (pantoprazole) to minimal (lansoprazole). They do, however, share VOD and protein binding properties. As a group, they are all highly (97%) protein bound, mainly to albumin, which restricts their volume of distribution to wherever albumin is found (i.e. extracellular fluid, about 0.2-0.3L/kg). Even more interestingly, the pyridine moiety of PPIs is protonated in the acidic environment of the parietal cells, effectively trapping the drug there. They are all weak bases with a pKa of around 3.9-4.5, which means they tend to accumulate in the highly acidic luminal environment of the parietal cell's secretory canaliculi, where the pH can be as low as 1.0. According to Shin & Sachs (2008), their concentration in those canaliculi can be 1000 times higher than their blood concentration.

Sucralfate is basically an inert chemical paste, and distributes nowhere except the gut lumen, where it remains awkwardly after all the other drugs had left. 

Misoprostol has a wide volume of distribution (about 16L/kg) and is about 90% protein bound, without any other exciting properties.

Metabolism

Antacids, for the most, undergo no biotransformation other than the chemical reaction which takes place as part of their normal function. The citrate of sodium citrate may also get absorbed to some degree, and be metabolised in literally the next Krebs cycle it comes across, turning into water and CO₂. The inorganic ions that were conjugated with the citrate or carbonate will obviously undergo no metabolism whatsoever.

Atropine is somewhat metabolised by the liver, through various steps involving glucouronidation and enzymatic hydrolysis. Of the total dose, perhaps 50% will undergo this metabolism, and the rest will escape metabolism entirely and exit via the urine, unharmed. According to Van der Meer (1986), the main barrier of researching its metabolism in man has been the toxicity of large doses, making it difficult to recruit willing volunteers for pharmacokinetic research. For their own study, the investigators were only able to convince a single normal subject to take a 2mg dose. From their urine, five different metabolites were recovered, including noratropine, tropine, tropic acid and atropine-N-oxide (suggesting numerous possible metabolic pathways). 

H2 receptor blockers are generally not metabolised very much, on average 20-50% of the administered dose undergoes hepatic biotransformation. Apart from cimetidine, which is of some limited borderline interest because of its tendency to inhibit CYP45 enzymes, the rest of them are relatively boring from a metabolic point of view. 

Proton pump inhibitors undergo hepatic metabolism by CYP450 enzymes, predominantly by CYP2C19. Some of them also inhibit CYP450 enzymes, which results in an improvement in their bioavailability over time. This is mildly interesting. A much more interesting biotransformation step takes place inside the parietal cells. Inside the extremely acidic environment of the parietal cells, PPI molecules undergo two protonation reactions, the first of which traps them in the cells, and the second (protonation of the benzimidazole or imidazopyridine ring) forms a sulfenamide derivative, which is the actual active form of the drug (so, in that sense, they are all actually pro-drugs).

Sucralfate undergoes no metabolism. Theoretically, what happens to it in the stomach could be described as "biotransformation" of a sort, as under the effects of gastric acid it undergoes a change from a complex salt to a polyanionic gel. In that sense, it could be described as a pro-drug of sorts. Upon making contact with acid and dissociating in the stomach, the oral suspension releases its aluminium (which goes on to do good buffering work) and becomes incredibly negatively charged (have a look at the sucralfate molecule, there's seven aluminium hydroxide groups hanging off it).  The resulting gelatinous mass of sulfated sucrose molecules binds avidly to anything and everything with a positive charge, be it gastric mucus, protein, peptide, drug, or lunch. 

Misoprostol really only has one exciting pharmacokinetic property, which is that metabolism rapidly eliminates the parent drug from the circulation and turns it into misoprostol acid, which has pharmacological activity. The parent drug is almost 100% metabolised by the liver.

Elimination

Antacids undergo two main mechanisms of elimination. Whatever antacid material that remains unreacted in the gut will often be eliminated in the faeces, producing some degree of osmotic diarrhoea. The absorbed inorganic ions will be eliminated according to the normal pathways which they would follow (i.e. renally). Even the ions which the kidneys are not usually expected to handle (eg. aluminium and bismuth) are still renally eliminated.

Atropine is partly eliminated as unchanged drug; 50% of it is recovered in the urine. The rest is metabolites. Its half-life is 2-5 hours, but the duration of effect on various organs could differ, as it has a tendency to lodge there (eg. its mydriatic effect in the pigmented eye could be as long as 96 hours).

H2 receptor blockers rely mainly on renal clearance in renally normal people, with as much as 50-80% of an administered dose excreted unchanged via the kidneys. Though they all have short half-lies (1-3 hours), the duration of effect is longer lasting, and most of them require only twice daily administration

Proton pump inhibitors are metabolised in the liver and end up being eliminated renally as inactive metabolites. Their half-lives are short, if you look at just the blood concentrations. This does not reflect the duration of their effect. The molecules trapped in the acidic canaliculi of the parietal cells are somewhat safe from liver enzymes, and the duration of their effect is sustained, allowing for daily dosing in the majority of cases. For most of these drugs, the recovery of proton pump function is mainly dependent on the production and luminal expression of new pump proteins, which is obviously not a very quick process, and so one can expect at least 24 hours of effect.

Sucralfate is eliminated reluctantly by the gastrointestinal tract, to which it clings for some sustained period. After 48 hours, most of the gel would have made its way out.

Misoprostol acid, the active metabolite of misoprostol, is eliminated over 2-4 hours and is cleared mainly by the kidneys (80%). The half-life of the parent drug is rather short (20-40 minutes), nor does the metabolite hang around for long, meaning it needs to be administered several times a day for its anti-ulcer effects.

Pharmacodynamics of drugs which influence gastric pH

Antacids neutralise gastric pH by a purely chemical mechanism. They are a largely alkaline bunch of simple inorganic chemicals which, in the presence of hydrochloric acid, produce some kind of a chloride salt and water. Considering the rate of acid production in the stomach, one might expect this to be an insurmountable Sisyphean task, and indeed it is said that a clinically significant suppression of gastric pH usually requires about 400 mmol/day of acid neutralising capacity (Ching & Lam, 1994).  This class of drugs is serenaded by Lam (1988) in an excellent retrospective, which is unfortunately only available to the owners of Baillière's Clinical Gastroenterology. They are also one of the most ancient drug classes, with various sources attributing the first use of "neutralising earths" to Celsus, an early Hellenic philosopher and vociferous critic of Christianity. It is impossible to determine exactly what the composition of these "earths" was, but they probably contained calcium carbonate, which is naturally found in chalk limestone and marble (thus being abundantly available to a Greek philosopher). 

Since the days of Celsus antacid science has moved only a fraction away from eating actual dirt. "Earths" have been replaced by purified (largely inorganic) compounds which are still basically mineral alkali molecules. As one might imagine, there's a vast variety of potential candidates for that role, as there are about ninety-five metal elements in the periodic table which could form carbonates or hydroxides and could therefore theoretically be used to decrease gastric pH. Of course, practically, nobody is going to prescribe uranium hydroxide to their patients, and so the list is somewhat limited to alkaline substances which are without any serious biological effect, besides their alkalinising properties. Even more practically, most people would agree that it would be unreasonable to expect these drugs to be completely pharmacodynamically inert. The situation in the modern era has not changed much since the times of  Lloyd Adams, who in 1939 wrote that  "unexpected difficulties are often encountered in the use of antacids to lower gastric acidity, because no antacid has yet been found which will simply neutralize acid without exhibiting additional pharmacologic action." With that, here is a list of common antacids in modern use, with their side-effects listed:

Antacid	Side effects
Aluminium hydroxide (Gastrogel)	Hypophosphataemia Constipation Aluminium toxicity
Calcium carbonate (Alka-Seltzer, Tums)	Hypercalcemia Metabolic alkalosis
Magnesium hydroxide (Milk of Magnesia)	Hypermagnesemia Osmotic diarrhoea Metabolic alkalosis (milk alkali syndrome)
Sodium bicarbonate (Gaviscon)	Hypernatremia
Bismuth subsalicylate (Pepto-Bismol)	Black tongue, black or grey stools Bismuth neurotoxicity (mood changes, delirium) Salicylate toxicity

For the majority, the duration of action is modest. The drug becomes depleted of its alkalinising effect, the stomach empties, and more acid is secreted. The best case scenario is probably where sodium citrate is used preoperatively, where the effects of anaesthesia and fasting on the stomach emptying rate can be an advantage. Viegas et al (1982) reported three hours of reliable activity. 

Anticholinergic drugs have been around for an incredibly long time, though it is unclear whether the Oracle of Delphi was able to usefully comment on their acid suppressant effects while she was completely blasted on shrooms. Following from this comment, it will be intuitively apparent to most readers that the side effect profile would have made these drugs relatively unpopular even in the 1930s, when there was little else. Even newer more selective agents fell by the wayside. A review of their pharmacokinetics by Feldman (1984) was written as a means of introducing pirenzepine, a selective M1 antagonist you probably have not heard much about because its positive effects on acid secretion are largely outweighed by its negative effects on gastric emptying and gut motility. 

In short, the useful mechanism of action of anticholinergic drugs acting on gastric acid secretion can be summarised as follows:

• The vagus nerve stimulates gastric acid secretion by the parietal cells (which produce mainly acid) and G cells (which in turn produce gastrin, and indirectly influence parietal cells)
• This is a cholinergic mechanism
• When acetylcholine is released at the postganglionic nerve terminals in the stomach, it mainly affects M1 M3 and M5 receptors, which are Gq-protein-coupled receptors
• Cholinergic stimulation of M1 and M3 receptors uses intracellular calcium as a second messenger, whereas M5 receptors exert their action by decreasing the intracellular concentration of cAMP.
•  Calcium signalling is an important lever you can pull to increase gastric acid secretion (Chew et al, 1992), 
• Thus, inhibition of muscarinic acetylcholine receptors has the effect of decreasing intracellular calcium, and inhibiting gastric acid secretion

As mentioned above, the side effects of anticholinergic drugs had made them relatively unpopular, as they consisted of the familiar:

• Tachycardia
• Mydriasis
• Loss of sweating and impairment of thermoregulation
• Delirium and hallucinations
• Constipation, ileus and urinary retention

To name but a few.

Histamine (H2) receptor antagonists block the effect of histamine on gastric acid production by antagonising the basolateral H2 receptor, and therefore decreasing the levels of cAMP in parietal cells. That cAMP is usually responsible for the activation of protein kinase A, which in turn phosphorylates all sorts of cytoskeletal machinery to bring H⁺/K⁺ ATPase transporters ("proton pumps") to the luminal surface. Ergo, the mechanism of action of these drugs is to reduce the expression of the acid secretion machinery. From this it follows that these drugs can never be as effective as the proton pump inhibitors, which block the pumps directly (Lamers, 1996).

Apart from their desirable effects, H2 receptor antagonists are largely without major side-effects (apart from the occasional carcinogenic excipient). Their main non-H2 effects are mainly related to pharmacokinetics; for example ranitidine famously inhibits alcohol dehydrogenase, and cimetidine interferes with CYP450.

Proton pump inhibitors,  as the name suggests, inhibit the ATP-powered H⁺/K⁺ exchange protein which moves potassium out of the gut lumen. 

The PPIs, after they are biotransformed into the chemically reactive sulfenamide form, bind covalently to these pumps. This alters the pump in such an irreversible way, that it can pump no more. This mainly disables active pumps expressed at the luminal surface, which does not disable all of them, as some are still in the cytosol waiting to be expressed (Strand et al, 2017); this means that a PPI dose will actually be more effective after a meal, when more pumps are sent to work at the cell surface.

This results in the neutralisation of gastric pH. One does not usually expect to block every pump, and so the pH may remain relatively acidic with routine oral anti-GORD use (perhaps 3.0 or 4.0), but still nothing like its normal state. To give the reader some impression of what happens with routine ICU practice, when Laine et al (2008) looked at the gastric pH of patients with bleeding ulcers receiving continuous infusions of intravenous PPI (lansoprazole), they found the pH remained above 6.0 for the majority of the 24-hour study period. 

Is there any difference between them? For the intensivist, probably not, as all have a relatively similar effect on the suppression of gastric acid secretion. For the healing of erosive oesophagitis, the therapeutic target is said to be a pH 4.0 or greater for more than 16 hours of the day, and most PPIs are able to effortlessly achieve this. In terms of potency and efficacy, pantoprazole seems to achieve a higher pH for longer (Shin et al, 2017), but there does not seem to be any data to support the use of one PPI over another for stress ulcer prophylaxis in the ICU. 

The side effects of PPIs do need to be mentioned, as even though the number needed to harm is fairly high, the fact that many ICU patients are prescribed a PPI makes them a vulnerable population, susceptible to these complications. The more notable effects (from Fossmark et al, 2019) include:

• Hypomagnesemia
• Interstitial nephritis
• B12 and iron deficiency
• Intestinal bacterial overgrowth
• Increased risk of C.difficile infection
• Increased risk of hospital-acquired pneumonia

Sucralfate, as already mentioned, becomes a highly reactive anionic sucrose polymer, binding to every positively charged residue and crosslinking itself with other molecules in the stomach. The effect of this is a high affinity binding to all mucosal surfaces, both eroded and not. Additionally, a fair amount of aluminium hydroxide is released by its activation, which is a respectable antacid in its own right.

The upshot of this gelatinous oversmearing of gastric mucosa is the formation of a protective barrier, which is thought to defend against erosion and protect already eroded areas from gastric acid. The rationale for this sort of therapy may be that these erosions need time to heal, and sucralfate can protect them until their normal mucosal integrity is restored. Theoretically, gastric pH can remain low with this treatment, which means that some of the objectionable consequences of pH-centric therapy are avoided (i.e. the acid should still be there to protect you from all the C.difficile and ventilator-associated pneumonia). The duration of the effect is somewhat unpredictable, as it persists for as long as the paste sticks to the ulcerated mucosa, and this mechanical feat is dependent on a lot of different factors. For most normal people who have the intention of eating solid food, sucralfate administration needs to happen at least every six hours for the maximum effect to be sustained (each dose is 1g).

Misoprostol and its active metabolite bind to the PGE2 class of G-protein coupled receptors, which come in four distinct flavours (EP1 through to EP4). For the purpose of antagonising parietal cell acid production, the EP3 variant is the most important one, which are G_i-coupled (Jones, 2007). Prostaglandin E2 usually acts as an antagonist of gastric acid production, which is the mechanism for NSAID-induced gastric ulceration (Musumba et al, 2009); therefore, to administer a potent PGE2 receptor agonist would also be expected to have the same effect. The intracellular messaging seems to be mainly related to the decrease in cAMP as the result of receptor binding, which reduces the expression of the proton pumps on the luminal surface. The attentive reader will correctly point out that this is a similar effect to that of the H2 antagonists. The two main side effects are diarrhoea and increased uterine contractility, both of which are usually undesirable.