Composition, volumes and regulation of gastrointestinal secretions

This chapter has some vague relevance for Section O1(i) from the 2017 CICM Primary Syllabus, which asks the exam candidate to "describe the composition, volumes and regulation of gastrointestinal secretions".  Fortunately for the exam candidates, this topic has been largely ignored by the CICM examiners, who have only used it in two SAQs:

• Question 16 from the second paper of 2020
  (formation/regulation of gastric acid)
• Question 7(p.2) from the first paper of 2009
  (functions of gastric secretions)

So, really, when the syllabus document says "gastrointestinal secretions", what they actually mean is "acid". Which is predictable and totally understandable, as the risk of stress ulceration and the use of PPI in ICU is an important topic, whereas saliva is not. Thus, emphasis here is on the mechanisms of acid secretion, and how it is controlled. Hopefully future exam questions remain faithful to the theme of acid instead of branching out into the other secretions. The full scale of this huge topic could easily overwhelm the emotional and physical reserves of an exam candidate if it appears late in the paper (last couple of questions, just as you're running out of time). 

Unsurprisingly, this unsexy subject matter receives little attention in the literature. Thomas (2006) is probably the single best overview, even though his paper somehow got published without any references whatsoever.  The content of that article is probably what the maximum amount of knowledge of this topic looks like, i.e. this is what the ICU trainees could reasonably be expected to carry into the exam. It represents the outermost edge of a massive and terrifying structure that nobody among the CICM trainees should feel compelled to explore, but if they really want to, Welcome (2018) will be the best-researched and most thorough reference, with over a hundred pages of extremely dense information. 

Volume of gastrointestinal secretions

Usually, textbooks that provide any detail about GI secretions produce a table where some kind of "daily fluid flux" through the gastrointestinal tract is presented, in terms of daily fluid flow. It usually looks like this one (from Camilleri, 2004):

Intake	Volume
Oral intake of fluid	1500ml
Saliva	500-1500ml
Gastric secretions	1000-2000ml
Bile	1000ml
Pancreatic secretions	1000ml
Small bowel secretions	3000ml

 The numbers differ by 500ml or so, depending on which textbook you pick, and they are not important. Somebody at some stage must have measured these and published on it, but because the values will vary so widely between individuals, the exact figure is unimportant. All you need to understand is that all of these 7 - 9 litres of fluid end up being reabsorbed, mainly by the ileum and colon. Only about 1200ml per day reaches the colon, in fact. In case you're wondering, the maximum amount of water the colon can absorb is apparently about 5700ml, according to some horrible-sounding experiments by Debongnie & Phillips (1971)

Oral secretions

"Saliva is a most valuable oral fluid that often is taken for granted", admonish Humphrey & Williamson (2001) in the opening statements of their comprehensive review. This is the product of minor (10%) and major (90%) salivary glands, which differ slightly in the composition of their secretions. The pedant would have to refer to it as a "mucoserous exocrine secretion". 

Composition: This is not an ultrafiltrate of plasma; it is a hypotonic fluid from which the salivary ducts extract most of the sodium (the theory is that this is because it enhances the hydration of protective oral mucin glycoproteins). The tonicity and composition changes with flow rate, which makes it difficult to offer an authoritative list of ingredients (but people have tried); generally even experts charged with the task of reporting on saliva composition tend to limit themselves to a short list of electrolytes.  The macromolecular components are multifunctional, redundant and amphifunctional (i.e. they all have multiple similar functions which could be simultaneously protective and detrimental), the complexity of which bedevils the makers of artificial saliva. Despite this appearance of complexity, all these extra protein molecules contribute minimally to its overall volume: saliva is 99.5% water by weight, unlike plasma (which is 7-8% protein).

Function:  The functions of saliva are so numerous that to list them all would be a major undertaking. The reader who needs that list is referred to  Humphrey & Williamson (2001). In brief:

• Lubrication and protection (barrier function, mainly mucins)
• Buffering action and clearance (mainly bicarbonate)
• Maintenance of tooth integrity  (remineralisation of enamel by calcium and phosphate)
• Antibacterial activity (immunoglobulins, especially IgA)
• Taste and swallowing (lubricating the food bolus, acting as a chemically inert carrier fluid to dissolve molecules for taste sensation)
• Digestion (amylase in saliva begins to act on carbohydrates)

Volumes: This is obviously going to be highly individual, and most authors give a range (500-1500ml). Of this, 90% is "stimulated" flow, which occurs in response to eating and chewing. At rest, the flow is minimal (around 0.1ml/min, or 6ml/hr, or about 150ml/day), and it slows to essentially zero during sleep. This does have some ICU relevance: our intubated patients are not going to be eating, and therefore will have minimal salivary flow. For example, Dennessen et al (2001) invaded the ICU with dental hygienists and found that they couldn't even measure the composition of the saliva, as there was not enough to sample. The implications for oral mucosal health and the microbiome of the posterior pharynx are considerable, and this is relevant because that's the microbiome that is festering around the top of the endotracheal tube cuff, preparing to invade the lung and cause VAP.

Regulation: The secretion of saliva is regulated by medullary nuclei in response to mechanical, gustatory and olfactory stimuli. The glands are innervated by both sympathetic and parasympathetic nerves, of which the latter seems to be the dominant force. Anticholinergic stimuli have a profound effect of reducing salivary flow, whereas the effects of sympathomimetic stimuli are fairly unimpressive (Ekström, 1989).

Oesophageal secretions

Yes, the oesophagus secretes stuff. No, it's not the stuff of legend. No epic sagas have been written to glorify this material. The oesophageal glands mainly secrete a thick bicarbonate-rich mucus which serves as an anti-reflux barrier. It is a reasonable amount of bicarbonate for an organ we often think of as just a gormless food tube. Meyers & Orlando (1992) were able to isolate ten-centimetre lengths of the oesophagi (oesophaguses?) of healthy volunteers by using two balloons, and measured a bicarbonate secretion rate of about 160 mmol/hr.

Gastric secretions

The stomach wall features glands that form deep tubular pits, lined with secretory cells. These cells all produce something slightly different, and it would probably be worthwhile to discuss them and their specific functions first before moving on to the content of the secretions and specifically the acid.

• Goblet cells secrete mainly mucus
• Chief cells secrete mainly pepsinogen and gastric lipase. Pepsinogen is converted into active pepsin in the low pH environment of the stomach.
• Parietal cells produce mainly hydrochloric acid and intrinsic factor
• Enterochromaffin cells secrete the histamine 
• G-cells secrete gastrin
• δ-cells or D cells  secrete somatostatin

All this gets mixed up in the tube pits, and comes out of the glands as a combination product. The glands are distributed unevenly along the anatomical areas of the stomach, and the cells are distributed regionally within each gland, such that:

• Glands in the cardia mainly produce mucus, and are dominated by goblet cells
• Glands in the fundus are called oxyntic as their main product is hydrochloric acid
• Glands in the pylorus tend to mainly produce gastrin

Composition of gastric secretions

• Water - when taken as a homogenised aspirate, gastric secretions are 95% water by weight - more water than blood, but less watery than saliva, and much thicker than you'd expect from that protein content because of the gel-like qualities of its mucin.
• Hydrochloric acid, which brings the pH of the mixture to a range of 1.5-3.5
• Mucus, which is rich in bicarbonate, capable of neutralising the aforementioned acid to protect the gastric epithelium. It forms an  "unstirred layer" on the walls of the stomach, which has about 20mg/ml of viscous mucin polymer (Kaufmann, 1981). This makes it thick enough to stick to the walls, keeping the central lumen pH in the desirably acidic range.
• Digestive enzymes:
  • Pepsin (90% of the total protein content)
  • Gastric lipase
  • Gelatinase
  • Intrinsic factor
• Electrolytes the concentration of which obviously quite variable, but occasionally textbooks will come out with a table of electrolyte concentrations that looks a bit like this:

  Electrolyte concentration	(mmol/L)
  Chloride	130-180
  Sodium	10-60
  Potassium	12-20
  Calcium	1.00-1.30

  This actually seems to come either from an ancient cat experiment by Gamble & McIver (1928), or from a still fairly ancient human experiment by Strong et al (1960). Every textbook will give a slightly different set of numbers, because the concentration strongly depends on the flow rate of the secretions and on whether or not the subject is being fed. The original data from Strong et al demonstrate just how variable this can be: in one subject, the sodium concentration fluctuated between 6 and 76 mmol/L, with a chloride concentration as high as 171 mmol/L.

Now, that's just the gloop secreted into the stomach lumen. From the character of the college answer to Question 7(p.2) from the first paper of 2009, it seems like when the college say "describe the functions of the gastric secretions", they meant all the secretions, including the endocrine or paracrine hormones. That's probably not how the candidates would have interpreted it, which is why 60% of them failed that question. Anyway, the specific hormones mentioned by the examiners were:

• Gastrin, produced by G-cells, which stimulates gastric acid secretion
• Histamine, produced by enterochromaffin cells, which also stimulates acid secretion
• Somatostatin, produced by δ-cells, which inhibits gastric acid production

Of course that's not all of them, as gut hormones are numerous. The only other one perhaps worth mentioning is ghrelin, a peptide hormone produced by epithelial cells of the stomach which influences satiety and appetite. Instead of going into excessive detail here, the reader can be left with links to comprehensive reviews by Gribble et al (2018) or Ahmed & Ahmed (2019) to fill in all the blanks.

Function of gastric secretions

You could probably just classify the main functions of exocrine gastric secretions into two main categories: digestive and protective. 

• Immune function:
  • pH acts to decontaminate bacteria in food. Martinsen et al (2019) had found numerous examples of increased risk of infection if this protective function were interrupted.
  • Proteolytic enzymes also act on microbial cell wall components and endotoxin
• Barrier functions
  • Gastric mucus acts as a barrier and neutralising agent against the gastric acid. This is a phylogenetically ancient function, probably developed in cartilaginous fish over 400 million years ago. 
• Macronutrient digestive function:
  • Gastric acid 
  • Proteolytic enzymes begin the digestion of the protein in the food bolus (pepsin is the main one, and it accounts for about 15% of the total protein breakdown according to Smith et al, 2010)
  • Gastric lipase hydrolyzes the ester bonds of triglycerides, contributing something like 30% of the total lipid catabolism in the gut (Gargouri et al, 1986).
• Micronutrient digestive function:
  • Pepsin also helps ferric iron (Fe³⁺) conversion to the more soluble ferrous (Fe²⁺) ion, according to some widely cited experiments from Jacobs & Miles, 1969.
  • Intrinsic factor binds to B12, creating a complex which can be absorbed in the terminal ileum
• Endocrine and paracrine functions
  • The regulatory effects of gastrin, somatostatin and histamine affect the rate of gastric acid secretion

For the intensivist, there is probably some hidden relevance here, as we are usually the people who a) disable the gastric acid secretion mechanisms, or b) bypass the stomach altogether by feeding patients nasojejunally. So:

• Clostridium difficile infection: altered microbiome of the upper intestine due to the delivery of normally alkaline chyme
• Gastrointestinal tract pathogens: the lack of normal acid in the gastric content could theoretically lead to the delivery of contaminated material to the distal gut, which could give rise to  all manner of enteritis (Campylobacter and whatnot) but this is less likely as ICU patients are usually fed a pre-mixed liquid diet which should be free from bacterial contamination.
• Malabsorption: A lack of gastric acid involvement in the processing of food (with nasojejunal feeding) will result in the malabsorption of nutrients, particularly of lipid (without the extra gastric lipase) and of B12 (as intrinsic factor will be missing)

Volume of gastric secretions

As mentioned above, this is fairly random and feeding-dependent, but at least it varies less than saliva in the critically ill patient. Most textbooks and plain language reviews will confidently state that the stomach secretes a constant 1000-2000ml/day of mucus and acid, though these data appear to be largely derived from horrific pre-War dog studies. The exact volume does not matter overmuch. For the intensivist, the rate of gastric secretion production is important mainly by the way it influences the safety if intubation, and by what it says about gastric emptying:

• Gastric residual volume in the ICU patient is a measure of feed tolerance and is usually assessed at four-hourly or six-hourly intervals; the volume of secretions aspirated from the nasogastric tube is measured. If the normal rate of secretion is about 100ml/hr, and then you add 60ml/hr of nasogastric feeds, and if no gastric emptying is taking place, the stomach will contain 640ml of fluid at the end of four hours, or 960ml at the end of six hours. So, the presence of only 500ml (or less) of fluid after a six-hour period is therefore considered a safe limit for "feed intolerance", as it means that at least half of the gastric content is making its way into the rest of the intestinal tract.
• Fasting before anaesthetic is to some extent influenced by the secretion rate, as aspiration during intubation is going to be a bigger problem when the stomach is full. Ong et al (1978) determined that the gastric aspirate volume of fasting outpatients was only about 70ml on average.

Regulation of gastric secretions

Schubert (2015) is probably the best reference for this, as his article organises this topic by hormone. The alternative method would be to organise the discussion by effect (eg. "acid secretion" and "mucus secretion", or by phase of digestion (cephalic, gastric, intestinal). Some combination of these systems was attempted here:

• Factors that increase gastric acid secretion: 
  • Parasympathetic nervous system (via the vagus nerve) - mainly during the cephalic phase of digestion, stimulated by the food bolus entering the mouth; mediated by acetylcholine, via the muscarinic M3 receptors. 
  • Gastrin, during the cephalic and gastric phases of digestion; released due to central signals as well as due to the mechanical effects of food in the stomach. Mediated by binding to gastrin receptors on parietal cells. Gastrin also binds to CCK-B receptors on enterochromaffin cells and causes them to release histamine.
  • Histamine; during the cephalic and gastric phases of digestion, released due to central signals as well as stimulated by gastrin. Binds to H2 receptors on parietal cells to mediate the release of acid
  • Mechanical stretch (by vago-vagal and local smooth muscle stretch reflexes)
  • Peptides, caffeine, alcohol - sensed by mucosal chemoreceptors of the antral G cells, which then release gastrin
• Factors that decrease gastric acid secretion
  • Somatostatin, if the luminal pH decreases to below 1.5
  • Small intestine hormones, during the intestinal phase of digestion (as the stomach empties into the duodenum):
    • Vasoactive intestinal polypeptide (VIP)
    • Secretin
    • Cholecystokinin
  • Prostaglandins PGE2 and PGI2, which also promote increased bicarbonate and mucus secretion in the gastric mucosa - these are the products of COX-1 activity, the loss of which is responsible for gastric ulceration seen with NSAID use
• Mucus secretion by goblet cells:
  • Mainly the result of some kind of mechanotransduction triggers, such as the distension of the stomach wall with food. 

Formation of gastric acid

Question 16 from the second paper of 2020 specifically wanted to focus on the mechanisms of acid secretion by parietal cells. Mainly to help himself understand what's happening here, the author had succumbed to the urge to illustrate this process using a primitive channel diagram. 

Acknowledging that this is not the right way to explain anything to anybody, a wordy version is also offered, as well as the obligatory peer-review reference in the form of Engevik (2020).

• Parietal cell structure:
  • The apical surface of the parietal cell has numerous small secretory canaliculi which increase the apical surface area of the cells
  • Resting parietal cells have numerous vesicles the membranes of which have acid-secretory transmembrane proteins
  • With stimulation, these vesicles fuse with the apical canaliculi to rapidly increase the secretory capacity of these cells
• Basal membrane ion transport in parietal cells
  • CO₂ and water are able to diffuse into parietal cells passively
  • Carbonic anhydrase converts the CO₂ and H₂O into HCO₃^- and H⁺
  • HCO₃^-  is then exchanged for chloride at the basal membrane
• Apical membrane ion transport in parietal cells
  • H⁺ generated by carbonic anhydrase is exported by an ATP-powered H⁺/K⁺exchange pump, otherwise known as the "proton pump" - this is the molecular drug target of PPIs 
  • The potassium used in this exchange is returned to the gut lumen by KCNQ1/KCNE potassium channels in the apical membrane of the parietal cells.
  • Chloride is exported through apical chloride channels (ClC-2)
• Net effect of parietal cell activity
  • In the gut lumen:
    • Chloride concentration increases from 120 mmol/L up to 180 mmol/L
    • pH decreases - most textbooks quote 1.5 as the lowermost figure, but in actual fact that is the median pH. Most studies report lower bottom figures. For example,  Ayazi et al (2009) reported a range of 0.3-2.9, with a median of 1.7
  • In the portal circulation:
    • Carbonic anhydrase activity consumes CO₂, which is therefore lower in the portal venous blood than in the arterial blood delivered to the stomach, i.e. the actively secreting stomach has a negative respiratory quotient
    • This increases the pH of portal venous blood during the cephalic and gastric phases of digestion, a phenomenon known as the "alkaline tide"
    • As the result, the systemic body pH increases transiently following a meal
• Influences that stimulate acid secretion by the parietal cells during the cephalic and gastric phases of digestion:
  • Acetylcholine (M3 receptors), using calcium as an intracellular second messenger, during the cephalic phase
  • Gastrin (gastrin and CCK-B receptors), via cAMP as second messenger, during the cephalic and gastric phase
  • Histamine (H2 receptors)  via cAMP as second messenger, during the cephalic and gastric phase
• Inhibitory influences mainly take effect during the intestinal phase, and include:
  • Somatostatin
  • Vasoactive intestinal polypeptide (VIP)
  • Secretin
  • Cholecystokinin

The secretion of bile

Without getting carried away too much with bile, as it is treated with dignity in the liver section, the following brief comments can be made:

Composition, in summary, is an alkaline (pH ~ 7-8) and hyperosmolar (~600 mOsm/kg) soup, containing:

• 95% water
• Bile acids (about 50-80% of the total organic molecules), which are referred to as "bile salts" when they are conjugated (usually with an amino acid, like glycine or taurine)
• Phospholipids
• Proteins/peptides such as glutathione
• Cholesterol
• Bilirubin

Function: 

• Emulsification of lipid droplets to facilitate their digestion by lipases
• Elimination route for waste cholesterol and bilirubin

Volume: About 500-600ml of bile is produced per day

Regulation can be separated into secretion and release:

• Secretion is mainly determined by the rate of bile salt excretion, which is in turn related to the rate of their de novo synthesis and their enterohepatic recirculation.
• Catecholamines, secretin, cholecystokinin, gastrin and glucagon all increase the volume of bile produced by the liver (Kaminsky & Nahrwold, 1971)  
• Release from the gallbladder is mainly mediated by cholecystokinin, which is released by the arrival of fatty acids into the small bowel

Pancreatic exocrine secretions

These come from the pancreatic acinar cells and ducts, which together comprise about 90% of the pancreas by volume. The acinar cells are responsible for the secretion of the digestive enzymes, and the ducts manage the electrolyte content. Owyang & Williams (2015) is the best review of this topic, in case anybody needs a level of detail much greater than what is necessary for CICM exams. 

Composition is rather interesting. This fluid is essentially iso-osmolar with the body fluid, and contain about the same amount of sodium and potassium as the plasma, but a vastly larger amount of bicarbonate. The chloride is used by duct cells to exchange for bicarbonate, producing a fluid that is highly alkaline, with a minimal chloride content (40 mmol/L) and an absurdly high bicarbonate content (upwards of 90 mmol/L at rest, and up to 140 mmol/L immediately after a meal, with a maximum pH of 8.3).

So alkaline are these secretions, that their loss (eg. by a pancreaticoduodenal fistula) leads to metabolic acidosis. Alternatively, looking at things from a Stewart-like quantitative physicochemical perspective, the acidosis develops because of the loss of a strong cation (sodium) and the reabsorption of a strong anion (chloride), all of which leads to a decreased strong ion difference in the extracellular fluid, and therefore acidosis. This does have some relevance to the ICU trainee, as the uncontrolled leakage of pancreatic juice or duodenal contents is an important differential diagnosis of a normal anion gap metabolic acidosis, a popular topic for exam questions. 

Function of pancreatic secretions is predominantly digestive, and all the extra bicarbonate carried by this fluid is necessary to buffer gastric acid. The digestive function is carried out by the enzymes, of which the most important and notable ones are  trypsin, lipase, and amylase

Volume of pancreatic secretion is usually said to be about 2500ml. 

Regulation is by a number of predictable endocrine and autonomic influences (Chandra & Liddle, 2009):

• Factors that increase pancreatic secretion:
  • Vagal stimulation and cholinergic drugs
    (including vago-vagal enteropancreatic reflexes)
  • Gastrin, cholecystokinin, and secretin
• Factors that decrease pancreatic secretion: 
  • Sympathetic stimulation
  • Somatostatin

Those vago-vagal enteropancreatic reflexes are probably the most important here. This is basically a reflex arc that is mediated entirely by the vagus nerve, including all the processing which takes place in the dorsal brainstem (Niebergall-Roth et al, 2006). Of the hormonal influences, the most important is thought to be cholecystokinin. The relevance of these regulatory mechanisms to the intensivist is probably minimal, except where they intersect with our interest in pancreatitis. In severe pancreatitis, one historical approach was to fast the patient for a prolonged period of time, as the concern was that enteric feeding would stimulate more autodigestive secretions. It has been largely debunked by findings that pancreatitis only gets better with early nutrition, and the recommendation to commence early enteral nutrition has been incorporated into all the major guidelines (eg. ESPEN, 2020). The modern understanding seems to be that the necrotic pancreas, stewing in its own digestive juices, is hardly going to listen to normal neurohormonal secretory stimuli. 

Intestinal secretions

Most of the secretions in the intestine are the product of  Brunner's glands in the duodenum (described beautifully by Krause, 2000) with a minor contribution from the crypts of Lieberkühn in the ileum (remember that the job of the ileum is mainly to absorb things, not secrete them). The crypts secrete various digestive enzymes and they do have mucus-producing goblet cells but the net direction of movement is in, not out. The duodenal Brunner's glands are therefore responsible for most of the secretions in the intestine.

Composition of this fluid is mainly bicarbonate-rich mucus. The bicarbonate content of these secretions is high enough that the overall pH of duodenal content ends up being around 7.4-7.8. 

Function is mainly protective and buffering. The job of this mucus is to coat the duodenal epithelium and protect it by buffering the highly acidic chyme coming out of the stomach. 

Volume  is approximately 1000-2000 ml per day, though it is admittedly difficult to separate this from the secretions coming from the pancreas, as they empty directly into the duodenum.

Regulation is by the same basic mechanisms as gastric acid secretion, which makes logical sense as you would usually want both to be secreted at the same time. Specifically, the secretion of bicarbonate-rich duodenal mucus seems to be regulated by vagal cholinergic output (Odes et al, 1993)