Aperients, laxatives and prokinetics

This chapter is relevant to Section O2(i) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "Describe the pharmacology of aperients, laxatives and other drugs that affect gastrointestinal motility". The CICM curriculum designers have tagged this subject with an L1 level of importance, which suggests we should know it in detail. This is counterfactual, as this topic has never appeared in the written exam, and so could not possibly be a priority for anybody. Well, no, to be fair, 30% of an ammonia question was related to lactulose, but really this is hardly any representation. What follows is an attempt to reconcile the daily clinical importance of the material with the need to focus on exam essentials. 

What could they possibly ask about these drugs in a foundational critical care exam? Certainly practical uses and applications would be left for later stages of training (eg. where the management of constipation in ICU or prokinetic agents and feed intolerance are discussed). Probably nothing about the ADME either, as most of these drugs really do not have any curious pharmacokinetic properties. The most likely question on this topic would probably be something about classifying laxatives and comparing their mechanisms of action and side effects. Thus:

This unglamorous topic does not attract a lot of detailed literature attention. The best you can expect, from a single article, is some broad brushstrokes outlining the major classes and their most interesting properties. Of the articles that do this, Xing & Soffer (2001) is probably the best-referenced. Avila (2004) and Schiller (1999) are also reasonably detailed, in the sense that too much detail would be unreasonable. 

Definitions and classification

This area of pharmacology is for some reason infested with an abundance of euphemistic terms. Through the literature, one finds sanitising terms like "aperients", "laxatives", "purgatives" and "evacuants" being used to distract from the fact that these are drugs designed to generate an overabundance of liquid faeces. 

Whatever the purpose of this nomenclature might be, it is certainly not to increase clarity. For example, the meaning and origin of the term "prokinetics" seems easy enough to grasp intuitively, but what exactly is the difference between aperients and laxatives? Etymology is of no help here (because when is it ever); "aperient" is an archaic term used probably since the 1600s, a derivative of the Latin aperire,  "to open" (the same root as aperture), whereas "laxative" comes from laxare, "to loosen". The exact choice of words probably does not matter, as opening or loosening the bowel is certainly not how these drugs work.  Popular online dictionaries and complementary medicine propaganda seem to make some sort of distinction between them, suggesting that aperients are "mild laxatives", but a lot of professional literature does not seem to support this, suggesting that the terms are interchangeable. It's a question that the reader should regard as unimportant. The rest of this chapter uses "laxative" purely because the letter "x" in its name increases its objective coolness by 75%.

Whatever you call them, fortunately  at least the subclasses of this group seem to be divided on some sort of vaguely objective basis, even though the exact names of each class might vary from textbook to textbook:

• Aperients and laxatives:
  • Bulk-forming laxatives which increase the mass of stool (eg. psyllium)
  • Osmotic laxatives which increase the water content of stool by altering the osmotic gradient across the bowel wall (eg. lactulose)
  • Lubricants which aid the passage of stool mechanically (eg. microlax and paraffin)
  • Stool softeners which act mainly on surface tension to increase the penetration of water and fat into stool (eg. bisacodyl and docusate)
  • Stimulants which directly affect the myenteric nervous system (eg. senna and castor oil)
  • Secretagogues which increase the secretion of water or mucus into the gut to alter stool consistency (eg. linaclotide and lubiprostone)
• Prokinetics:
  • 5-HT agonists such as cisapride and tegaserod
  • Motilin agonists such as macrolides
  • Cholinergic agents such as neostigmine
  • Opioid antagonists such as naxolone, naloxegol, and methylnaltrexone

These classifications seem to enjoy a wide acceptance, and replace older terminology, which had a somewhat more dramatic character. To quote from "The Therapy of Dropsy" by Marvin (1940):

"The classical division of aperients is into - Laxatives, which increase the bulk of the stools but do not alter their consistency; purgatives, which produce unformed motions; hydragogues, which are followed by watery stools; drastic purgatives, which excite an inflammatory reaction in the intestine..."

There would probably be no practical benefit from a detailed exploration of each member of each class. Broad class-based generalisations will be discussed in the short summary which follows, with digressions on pharmacological properties taken only where the tangent seems valuable.

Bulk-forming laxatives

This group is difficult to define very precisely. Most authors mention that they increase the mass or volume of stool, which means that theoretically small metal spheres could be described as "bulk-forming". A better class definition would probably incorporate their main mechanism of action, which is to accumulate bowel water and make it inaccessible for reabsorption, thereby increasing the water content of stool.

Often described as "gentle" or "natural", bulk-forming laxatives are usually derived from some undigestable plant carbohydrate, and are usually long starch molecules. There are two main variants on this theme: branching molecules which end up being arranged randomly (producing a non-viscous colloid solution) and linear polymers which either arrange themselves regularly or which cross-link between adjacent fibres. The latter variety form water-trapping gels, often referred to as mucilage, as if this topic did not have enough weird gross terminology.

Of the most common ones:

• Psyllium, the powdered husk of Plantago sp. seeds, is one of the cross-linking carbohydrates which creates a gel. 
• Methylcellulose, a colloid laxative which does not form a viscous gel, is the product of cellulose after it is heated with sodium hydroxide and treated with methyl chloride. 
• Polycarbophil is the calcium salt of polyacrylic acid crosslinked with divinyl glycol, which does not dissolve in water, but which absorbs vast amounts of water (up to 100 times its own weight).
• Numerous other options (soy polysaccharide, pectin, etc) are available, and some, like hydrolysed guar gum, are already integrated into nasogastric liquid diet formulae.

Of these, all have basically the same properties, and similar side-effects:

• Abdominal distension: psyllium and methylcellulose may be metabolised by some gas-producing gut bacteria, giving rise to bloating and flatulence. Not so for polycarbophil, which is such an unnatural foreign substance that all organisms everywhere seem to completely ignore it. 
• Bowel obstruction: the conditions which are required for this are rare in the community, but they do occur (eg. Hefny et al, 2018). A lack of enteric fluid is often blamed, which would not be uncommon in the ICU.
• Delayed absorption of nutrients may result from the trapping of aforementioned nutrients within a gelatinous mass, from inside of which they would need to diffuse. 
• Delayed absorption of other drugs may occur by the same mechanism 

The mechanism of action of all these substances is basically the same, uniting them as a class. By ensuring that water remains in the stool, they increase the volume of intestinal content, thereby producing an increased distension of the intestinal walls. This stretch is a peristaltic stimulus, which then promotes gut motility. Following from this, one might rightly expect these drugs to lose their effectiveness in scenarios where intestinal water is lacking (eg. patients with stomach emptying), or where the intestine is not inclined to respond to normal signals (eg. where it is infected, bruised or ischaemic). In short, these may not be the best choice in the critically ill patient.

Osmotic laxatives

This group functions very much like bulk-forming laxatives, in the sense that all of these agents increase the water content and volume of stool. Like the bulk-forming laxatives, the osmotic group is characterised by being molecules resistant to absorption in the gut. Unlike the long and branching starchy molecules of the bulk-forming group, osmotic laxatives tend to be small molecules, and instead of trapping water in some kind of viscosity-enhancing hydrogel, they prevent its absorption by increasing the osmolality of intestinal content. One might expect there to be a lot of overlap between these groups, as psyllium and friends all also have an osmotic effect, but because those arabinose/xylose polymers have a molecular weight of about 1500 kDa, their concentration in the gut is usually measured in picomoles per litre (one mole = 1500 kg). In contrast, the molar mass of lactulose is 342g/mol, which means one's daily dose might be up to 500mmol - an osmotically relevant amount, considering all of those molecules plan to stay in the gut lumen.

Following from this, one might expect any molecule which is not absorbed in the gut to act as an osmotic laxative, and this is actually broadly very true. Basic nutrients can function as osmotic laxatives if the absorptive capacity of the gut is decreased (eg. norovirus) and this effect is a common cause of diarrhoea in the ICU, where osmotically active NG feeds are administered into damaged intestinal tracts. However there are several classes of molecule which we use as intentional osmotic laxatives, which are:

• Ionic salts, usually magnesium, as it is probably the most benign of all the electrolytes, and can therefore be administered in relatively large quantities, in the same way that for example potassium could not be. Sodium sulfate and phosphate are other, older alternatives. 
• Saccharides, such as lactulose (a disaccharide made up of glucose and fructose)
• Sugar alcohols, such a mannitol sorbitol and xylitol, which are the source of the warning on packets of mints, cautioning the eater that "excess consumption may have a laxative effect"
• Macrogols, i.e. different molecular masses of polyethylene glycol, which are the most common active ingredient of "bowel prep" used to evacuate the tenants of the gastrointestinal tract in preparation for colonoscopy.

The word "macrogol" is used to refer to the range of different polyethylene glycols which are commercially available. Just this one manufacturer seems to be selling different versions ranging from 400 Da to 8,000 Da. The macrogols in common medical use for bowel lavage seem to be composed of mainly PEG-3350, i.e. with a molecular mass of 3,350 g/mol. Apparently the cut-off is about 1500 Da; anything more than that is not absorbed, and therefore has a chance to make its mark on the colon.

Irrigating your gastrointestinal tract with hundreds of mmols of something unabsorbable cannot be without its consequences. The undesirable effects of these drugs are:

• Abdominal distension: sugar alcohols and lactulose can act as metabolic substrates for various gut bacteria, and this is in fact behind much of the beneficial effect of lactulose when it is used to treat hepatic encephalopathy. The unfortunate result of this is an overabundance of gas.
• Malabsorption of nutrients:  the promise of rapid gut transit brings with it the possibility that the time spent in contact with absorptive surfaces will be short. Absorption of drugs and nutrients may therefore be decreased.
• Electrolyte derangement: the diarrhoea produced by these laxatives will also produce electrolyte depletion, just like any other sort of diarrhoea.
• Volume depletion: your colon reabsorbs water for a reason, and to prevent this natural process robs the body of the ability to maintain its water balance through thirst. This should not be a major problem in the ICU, where IV fluid sources are plentiful, but it can certainly be a problem in the setting of community medicine.

Lubricant laxatives

Occasionally referred to as "emollient" laxatives, these substances exert a purely mechanical effect on the contents of the bowel, emulsifying the stool and making it more slippery, thereby decreasing the resistance to its passage. Or at least, this is the theory. The bowel produces enough mucus, and it would be surprising to consider that it might need some kind of help. 

• Mineral oil, a mixture of high-molecular-weight alkanes, is the only member of this class routinely available to the intensivist. It also goes by names like" liquid paraffin" and "paraffin oil". Alkanes are thought to be quite chemically inert, but prolonged exposure is said to be carcinogenic, which is why most normal people will not recognise this substance as a laxative. Administering it rectally has a similar effect to administering it orally, and in either situation the alkanes are not absorbed, at least where their carbon number exceeds 35.

Apart from being insidiously carcinogenic, hydrocarbon-based lubricant laxatives have all kinds of other side effects:

• Decreased absorption of fat-soluble vitamins: the abundance of unabsorbed nonpolar solvent in the stool keeps these vitamins from being absorbed normally
• Pruritis ani: as if your ICU patients need another reason to be uncomfortable and agitated
• Lipid pneumonitis as the consequence of aspiration, or from inhaling the vapour (though in all honesty one could simply purify the alkanes to the point where there is nothing left that might be volatile at normal room temperature)
• Awkward leakage, which is probably less of an issue in the ICU population, is a phenomenon which ruins the dinner party as the oily content of the bowel defeats the anal sphincter and leaks detectably on to the upholstery. This does not seem to be something serious enough to merit an appearance in medical case reports, but does seem to be serious enough for various patient-oriented literature to mention.

Stool softeners

Like the osmotic and bulk-forming laxatives, these drugs aim to increase the water content of stool; but they achieve this without as much interference with the total bowel water per se. Instead, they increase the access of the existing water to the stool, separating its particles and decreasing its viscosity by their surfactant effects.

• Docusate, dioctyl sulfosuccinate, is the best known example of this class. It was originally patented as a detergent, the properties of which were so remarkable that it appeared on the cover of Life magazine, and fifteen years later people were still recalling the picture (it was a duck, sinking helplessly into a tub of docusate-tainted water which was robbed of all surface tension).  

This drug is relatively safe, but still has some adverse effects:

• Increased absorption of alkanes: by acting as a surfactant, docusate makes all fats more available for absorption by increasing the surface area (decreasing the size and increasing the number of micelles). The result is that drugs such as liquid paraffin alkanes, formerly unabsorbable, become better absorbed. This is not very good: they remain inert, but now that they have gained access to the body, one has no mechanism of getting rid of them, and they form lipogranuloma inclusion bodies in the liver.
• Increased (more rapid) absorption of lipid-soluble drugs, by the same mechanism - which is not really a side effect if you intend for this to happen; and for that reason docusate is often included as an excipient to encourage rapid intestinal absorption.

Bowel stimulants

Bowel stimulants could equally be rebranded as "bowel irritants", "colon botherers" or "intestinal agitators". There are several chemical families, which occasionally include surfactants like docusate, perhaps because it is felt to have some sort of stimulant role.

• Diphenylmethane derivatives:
  • Phenolphthalein, familiar to many highschool chemistry students as that stuff that turns pink in the presence of alkali, was once a popular over-the-counter laxative, which has been withdrawn from the market because of a suspicion of carcinogenicity. It contains free diphenolic groups which are thought to be responsible for its bowel-angering properties.
  • Bisacodyl, a diacetic acid esther which is hydrolysed by endogenous esterases into aforementioned diphenolic groups which end up doing all the work. 
  • Sodium picosulfate, which is really just bisacodyl disulfate, is hydrolysed by bacterial arylsulfatase turning into the same active metabolites as bisacodyl (Godding, 1988).
• Fatty acids
  • Ricinoleic acid, the active metabolite of the triglyceride commonly known as castor oil, which appears to activate EP₃ prostanoid receptors (Tunaru et al, 2012). 
• Anthraquinones
  • Senna, commonly paired with docusate, is a crude mixture of plant materials which passes though the upper gut and ends up being hydrolysed into rhein monoanthrones in the colon (Godding, 1988).

These drugs act by a variety of mechanisms, which are not very well described in the literature. The best you tend to get is a paragraph or two, alluding to effects without describing their mechanism. For example, Godding (1988) mentions that polyphenolic drugs (eg. bisacodyl) act by increasing the permeability of the gut mucosa,  by inhibiting enterocyte Na⁺/K⁺ ATPase, and by increasing gut motility. Senna on the other hand induces the net secretion of fluid into the colon. How does it do that? Hands are waved in the direction of prostaglandins and serotonin. We can console ourselves with the thought that something like this could never possibly become the topic of an exam question.

Adverse effects are somewhat better documented:

• Pseudomelanosis coli: this is an effect mainly associated with the anthranoids, and is a direct toxic effect on the colon, where apoptosis of colonic epithelial cells occurs, and produces the deposition of dark lipofuschin granules. This weird pigmentation may or may not be linked to an increased risk of colorectal cancer.
• "Cathartic colon":  This condition may exist purely in the imagination of elderly gastroenterologists, and is based mainly on characteristic findings seen radiologically during barium enema studies. Specifically, one would see "radiographic changes of colonic redundancy and dilation with loss of haustral markings" (Xing & Soffer, 2001).
• Enteric neuropathy: a degeneration of the enteric nervous system which is clearly demonstrated on microscopy. This is also not clearly just a phenomenon related to the use of the laxatives- it may also be due to the condition which requires laxative use, eg. diabetic autonomic neuropathy or some sort of inflammatory bowel disease.

Secretagogue laxatives

These are relatively novel drugs which increase the secretion of water into the gut lumen, mainly by targeting chloride-mediated water uptake.

• Linaclotide is a 14-amino-acid peptide which acts on  guanylate cyclase-C. This results in an overabundance of cGMP, a secondary messenger which stimulates colonic chloride secretion. Chloride secretion in the colon leads to increased water loss into the lumen, and watery stools are the result.
• Lubiprostone is a bicyclic fatty acid which activates Type 2 chloride channels in the colon, with basically the same result (increased water secretion).

Though one might expect some metabolic alkalosis to develop as the result of all this chloride loss, in actual fact these drugs are generally well tolerated, presumably because one would be able to titrate the dose (profuse chloride-wasting watery diarrhoea would not be the desired endpoint).

5-HT₄ agonist prokinetics

This mechanism underlying the prokinetic effects of this group are well described in Gwynne & Bornstein (2019). In short, serotonin is an important signalling molecule in the gut, and in fact about 90% of the serotonin in the body is synthesised by mucosal enterochromaffin cells. Its main role appears to be as a neurotransmitter for myenteric neurons, and it is therefore involved in mediating the process of peristalsis, including the migratory motor complex and the myenteric reflex. Thus, to activate 5-HT₄ receptors means to stimulate peristalsis. Several agents are available to achieve this:

• Cisapride, a piperidinyl benzamide which has been withdrawn from the market because of cardiotoxicity 
• Tegaserod, an aminoguanidine indole derivative of serotonin with nonselective partial agonist activity against all the subclasses of the 5-HT₄ receptor, which has also been withdrawn from the market because of its adverse effects 
• Prucalopride, a benzofurancarboxamide derivative which has minimal activity at other serotonin receptors, and which is seen as a less toxic alternative to cisapride and tegaserod, and still hasn't been withdrawn from the market (it became available in 2018).

There are many relative substances, such as mosapride and velusetrag, which do not have any hideous adverse effects, and which are in routine use around the world.  It would be unfair to associate these newer safer drugs with the memorable adverse effects of the older drugs, which are:

• Ischaemic colitis (tegaserod)
• Cardiac ischaemia (tegaserod)
• QTc prolongation (cisapride)

Macrolide prokinetics

It is not the intention of this short section to cover the pharmacology of macrolides, as most of them are usually used as antibiotics. Thankfully, there is some literature (eg. Peeters, 1993) that focuses specifically on their prokinetic properties.

• Erythromycin is probably the best known representative of these, although basically all the macrolides have prokinetic activity. Clarithromycin is the other drug which has been used as a pro-kinetic in ICU, and azithromycin has been used to treat gastroparesis, as another example.  In practical terms, if one is looking after a critically ill patient and they are already on azithromycin for antibiotic reasons,  there is no merit in adding erythromycin for added prokinetic effect. The reason we tend to use erythromycin is mainly because its life cycle as an antibiotic has basically ended, with many community-acquired pathogens now exhibiting at least a modest resistance. Moreover we now have many safe alternatives, which means we no longer rely on it for its antibiotic effect. Lastly, erythromycin is used in doses which are much smaller than the effective bacterio-inhibitory doses, which means it should not produce resistance (as it would not apply a selection pressure on the microbial population).

The mechanism of prokinetic action of macrolides is related to their ability to act as motilin receptor agonists. Motilin is a peptide hormone which acts at multiple levels of the gut, from the antrum down (the only region which does not seem to respond is the ileum, which does not express motilin receptors). Activiating motilin receptors stimulates contraction by activation of smooth muscle L-type calcium channels, which manifests as increased antral activity, increased gastric emptying, and more high-amplitude propulsive contractions (Phase III of the migratory motor complex). If that sounds attractive, and one begins to wonder whether we really need to rely on the side-effects of antibiotics for these effects, one would be pleased to learn that there are motilin receptor agonists with no antibiotic activity, such as camincinal and mitemcinal (Sanger et al, 2012), some of which might eventually reach the market.

Macrolides are an old antibiotic class, and are not without their side-effects.

• QT interval prolongation is a feature of all macrolides
• Potential effects on antimicrobial resistance cannot be ruled out, as prescribing an antibiotic in subtherapeutic doses is exactly the sort of thing that can cause drug resistance, and theoretically the resistance to one macrolide can extend to other macrolides. This is explored in detail by Dall'Antonia et al (2006).
• C.difficile may become a problem, although the idea that sub-inhibitory erythromycin use directly causes this infection remains to be established
• Tachyphylaxis  tends to develop over weeks of therapy, as it would with any receptor system, i.e. motilin receptors become fewer in number and the effectiveness of the same dose of erythromycin diminishes. 

Cholinergic prokinetics

Acetylcholine is a major neurotransmitter for the intrinsic nervous system of the gut, and so it theoretically it would make some sense that we should look at using it as an instrument to reawaken colonic inertia. Practically, in the ICU, this use remains mainly in the realm of avoiding and preventing anticholinergic effects, rather than therapeutically using pro-cholinergic medications. The reason for this is that cholinergic prokinetics are too effective. For most intensivists, they represent the nuclear option. 

• Neostigmine, but also every other stigmine such as physostigmine or pyridostigmine are acetylcholinesterase inhibitors which will inhibit acetylcholinesterase everywhere, not only in the gut. 
• Itopride, which sounds like it should function as a 5-HT4 agonist, is also in fact an acetylcholinesterase inhibitor.
• Bethanechol, a direct muscarinic receptor agonist, was one of the oldest prokinetic agents (Megens et al, 1991), and is probably the only direct muscarinic agonist available on the market at the moment.

Acetylcholinesterase inhibitors increase the availability of synaptic acetylcholine, whereas the direct agonists just bind straight to the muscarinic receptors, but the mechanism of effect beyond this is basically the same. To reduce the articles by Sanders et al (2016) and Eglen (2001) into some easily remember points:

• Gastrointestinal smooth muscle has M2 and M3 muscarinic receptors
• M3 receptors are G_q/11 protein-coupled, and stimulation of these receptors therefore results in production of diacylglycerol and IP₃.
  • Diacylglycerol and ionised calcium activate protein kinase C isoforms
  • At the same calcium sensitisation occurs through the activation of Rho kinase.
  • These molecular pathways result in increased availability of intracellular calcium, and an increased sensitivity to calcium
  • The effect of this is increased contractility of the intestinal smooth muscle.
• M2 receptors are G_i-coupled and activation of these receptors results in the decrease of cAMP and reduced activity of protein kinase A.
  • The result is an opposition to the relaxation mediated by the neurohormonal receptors that increase the activity of those enzymes, such as β-adrenoceptors or 5-HT4 receptors
  • They also facilitate the opening of non-selective cation channels, which permit the entry of sodium, and therefore are synergistic to the release of calcium ions from intracellular stores
  • The overall effect is to potentiate M3-mediated contraction, and to counteract the relaxation-enhancing effects of other neuropeptides

Side effects? All of the effects are side effects. A reasonable person, reviewing the list of physiological reactions from the administration of these substances, would be forced to conclude that, whenever they are used for a specific indication, so many things inadvertently happen that the vast majority of the resulting outcomes are undesirable. The parasympathetic nervous system is vast and all-knowing, and to make it angry with a potent stimulant would obviously result in a million different effects. To name just a few, moving in the direction from head to toe,

• Miosis
• Salivation
• Increased bronchial secretions
• Bradycardia
• Increased gastric emptying
• Increased small intestinal and colonic motility
• Increased intestinal secretions
• Decreased detrusor tone
• Potential urinary and faecal incontinence

Opioid antagonist prokinetics

It would be amiss to leave these drugs out of the discussion, as they do often come in handy even though their mechanism of action is not "prokinetic" per se. They do increase motility, but only by restoring it to some normal level in scenarios where the motility was impaired by opioids. Townsend et al (2004) describes the mechanisms of this impairment in some detail. In summary, δ- and κ-opioid receptors seem to be expressed in the nerves of the myenteric plexus of the small and large intestine, and their activation produces a direct activation of smooth muscle cells, leading to a constant tonic contraction of smooth muscle, and this sustained contraction inhibits normal peristaltic activity. There does not appear to be any tachyphylaxis or tolerance for this effect, which means that even veteran opium fiends develop constipation.

To defeat this, two main agents are available:

• Naloxone, that old staple of opioid reversal, has been used as a prokinetic of sorts, where gut motility and feed intolerance in the critically ill were felt to be impaired mainly by the activity of opioids. In order to be effective in reversing gut stasis and still not too effective in reversing systemic analgesia, this drug is not given IV, but enterally - where it has minimal systemic bioavailability. Fortunately, in sufficient quantity, it does penetrate well enough into the intestinal wall, where it directly acts on the aforementioned myoenteric δ- and κ-opioid receptors. That sufficient quantity appears to be 5-8mg every 6 hours, which is about twenty times more than the usual IV dose.
• Methylnaltrexone and alvimopan are opioid antagonists which do not penetrate the blood-brain barrier, and which therefore should not produce any sort of reversal of desirable analgesic effects.

The side effects of these drugs are basically extensions of their desired effect:

• Counter-analgesic effect is possible, i.e. not entirely ruled out with either drug option, as there is a theoretical small fraction of enteral naloxone which may still make its way into the systemic circulation, and there is a theoretical penetration of methylnaltrexone into the CNS through a broken blood-brain barrier.