Endocrine functions of the pancreas

This chapter is probably at least slightly relevant to Section U1(i) from the 2017 CICM Primary Syllabus, which asks the exam candidate to  "describe the exocrine and endocrine functions of the pancreas". The exocrine function of the pancreas is explored well enough in a dedicated chapter from the gastrointestinal section, as it falls more neatly into Syllabus Section O1(i)), "describe the composition, volumes and regulation of gastrointestinal secretions". Moreover the endocrine and exocrine functions of the pancreas could not be any more different and unrelated, to the point where you'd be forgiven for thinking of it as two separate organs.

In summary:

The 2004 entry on the "Endocrine Pancreas" by Susan Bonner-Weir from the Encyclopedia of Gastroenterology is a good quick overview of the topic and contains more than enough detail to pass any exam questions about this aspect of pancreatic function. It is unfortunately paywalled by Elsevier. A free paper which covers the same ground but in a lot more detail is a review by Da Silva Xavier (2018). It is referred to extensively in the summary that follows, but it is probably too long for the trainee who just needs to skim an outline.  Thankfully, the expectations of CICM are clearly quite low, as this topic has rarely appeared in the exams - instead, more focused questions about the regulation of glucose tend to be asked.

Functional anatomy of the endocrine pancreas

The endocrine pancreas is really just a handful of glandular tissue scattered haphazardly through the pancreatic parenchyma. These secretory cells are really just a bunch of little tents pitched in the midst of the massive manufacturing complex of the exocrine pancreas, which weigh only about 1 gram in total (whereas the rest of the pancreas weighs probably about 70g in the adult male). Their hormonal products are ancient and highly preserved on an evolutionary level (even things without a recognisable digestive system seem to have some insulin-like peptides), but the arrangement seen in humans is somewhat unusual, as many other species have the endocrine pancreas as a separate organ. It is usually found close to the hepatobiliary system, which makes sense, as J.R Henderson had put it:

"If one were rash enough to design a pancreas the endocrine part would probably be a separate organ, occupying a position close to the gut in order to monitor the products of digestion being absorbed from the gut"

Invertebrates and primitive chordates like Amphioxus seem to have some endocrine insulin-secreting tissue scattered randomly around their intestinal tract. Hagfish have a endocrine pancreas which still lives on the surface of their gut wall, but is organised into a discrete lump, and in teleost fish it is a separate nodule sitting alone on the mesentery, well separate from the exocrine pancreas geographically. As you level your vertebrate attributes, your pancreases seem to fuse, to the point where reptiles seem to have a conjoint system (the same organ is host to a diffuse collection of endocrine and exocrine cells without any distinct islet formation). Birds and mammals have well-defined regions of endocrine cells which are clearly separated from their exocrine colleagues, a histological finding which Paul Langerhans first noticed in the rabbit.

How did this happen? How did this association form? The pancreas produces digestive enzymes which  help acquire macronutrient molecules for the circulation, and it produces regulatory hormones which help these molecules make their way into cells. Thus, the idea that at some distant evolutionary point those enzymes and hormones were the same molecule is attractive from a logical perspective, and some authors have proposed a hypothetical animal ancestor that secreted a "protoinsulin" digestive enzyme into their gut lumen, or used it as a neurohormonal coordinator of gut motility (Steiner & Chan, 1988). It appears that the need for blood glucose regulation arose at the same time as the need for a complex central nervous system arose, as the latter requires a stable supply of nutrient, and over the course of its development the insulin-producing cells transformed from something that purely regulated gut glucose concentrations to something that now also managed it in the bloodstream.

But if this were a conversation, the patient collocutor would by this point be making sustained efforts to redirect the rambling discussion back to functional anatomy. So:

Islets of Langerhans

The Islets of Langerhans are maybe a million or more tiny 50-500 μm blobs of endocrine tissue which are scattered through the pancreas, accounting for only 1-2% of its total mass. They are not scattered randomly - they are organised in clear "routes", elongated clusters of bunched islets, as this cadaveric study by Ionescu-Tirgoviste et al (2015) had revealed. Below, a slightly mangled reinterpretation of their original images illustrates the thick routes of islet-dense tissue coursing through the 3D model of a pancreas:

These long clusters surround some of the larger blood vessels, which makes sense, as their hormonal products need to be directed into the portal circulation. From head to tail, the islets are unequal in their cell population, with the islets of the tail and body being more involved in insulin production.

Cellular population of the islets of Langerhans

If a trainee is ever asked to list the main cellular populations of the endocrine pancreas, they would be expected to mention five main cell types:

• α-cells, which produce glucagon (30%)
• β-cells, which produce insulin (60%)
• γ-cells, which produce pancreatic polypeptide (5%)
• δ-cells, which produce somatostatin  (10%)
• ε-cells, which produce ghrelin (small fraction)

Realistically, you could probably omit the epsilon cells, and still pass. Also, it is unlikely that anybody would ever ask a student to describe or draw the structure of an islet, as their fine structure is pretty unimpressive. However there are a few main structural elements which might be of some interest:

• They are encapsulated,  with a one-fibroblast-thick wall separating them from the rest of the pancreatic tissue, and defining the islet interstitial space.
• They are well innervated with sympathetic and parasympathetic fibres. There do not appear to be any specialised synapses here, which probably means that neurotransmitters are released into the interstitial space to wash randomly over all the cellular residents in an islet.
• They are highly vascular: this 1-2% of pancreatic mass receives about 15% of the total pancreatic blood flow, and about 8-10% of each islet volume is dedicated to vascular tissue. The capillary network of each islet is similar to what one might see in the glomerulus, and the efferent capillaries of some of the smaller islets tend to pass through some exocrine tissue on their way back to the venules, giving the impression that some kind of portal system might be operating.
• Their capillaries are fenesterated. Notably so: some authors describe the postcapillary efferent vessels of some of the larger islets as being the leakiest of all capillaries, i.e their fenestrae are the largest, and the passage of large molecules through them is the easiest. It is though that the point of this feature is to allow the diffusion of secreted hormones into the bloodstream, and also into the surrounding pancreatic acinar tissue (in a paracrine manner). This is plausible, as pancreatic hormones are huge polypeptide molecules, and need all the help they can get to diffuse.
• The cells are organised around the capillaries, with β-cells forming the inner "core" around the afferent artieriole and capillaries, and other cells making up an outer "mantle" more associated with efferent vessels.

Unfortunately the author was helplessly attracted to the garish neon of immunofluorescence histology, and has caught himself reproducing some microphotography from El-Gohary & Gittes (2018), even as he acknowledges that the educational benefit to the reader is likely minimal. The images below capture a mouse islet stained with a green CD31 marker (a classic probe to define vascular endothelial tissue). Additionally, β-cells are glowing blue and α-cells are a bright red:

In case images do help, here is a great watercolour from the infancy of pancreatic histology by William Bloom (1931). The author used a then-novel Mallory-Heidenhain azan trichrome technique to stain each cell type a different colour. The α-cells are orange, δ-cells are blue and β-cells are a sort of light peach; central blood vessels with chunks of erythrocyte can be seen in the middle.

Insuloacinar portal system

The pancreas probably plays host to the insuloacinar (or pancreatico-acinar) portal circulatory system. One must say "probably" because it is not clear that this actually exists in humans. Those textbooks that do describe this forgotten culdesac of the circulatory system tend to say that it is a series of leaky portal capillaries that connect some of the smaller pancreatic islets to some of the acinar exocrine tissue. The objective of this system is said to be the carriage of growth factors and regulatory hormones from the islets to the  acini, and in support of this, people will point to the large heavily granulated cells of the acini which are adjacent to islets. Clearly those cells are being stimulated to grow larger, histologists exhort. The significance of this for the CICM exam candidate is unclear, but it is definitely not one of the important portal circulation systems which the college had expected trainees to mention in past exam papers, suggesting that one can safely ignore it. 

Now, for the hormonal products. Their treatment here will be fairly supericial, as the physiology of insulin, glucagon and somatostatin have their own dedicated chapter and attract their own set of CICM questions, while the other pancreatic hormone products are of zero relevance for CICM exams (as no examiner in their right mind is going to write a question about ghrelin).

Glucagon secretion by pancreatic α-cells

Glucagon is the hormone of the fasted state, and its main role is the prevention of hypoglycaemia by stimulated the release of glucose from hepatic and muscle glycogen stores. It is a 29-amino-acid peptide secreted by α-cells in response to regulatory influences which include sympathetic stimulation, intrinsic glucose sensing, and paracrine signals like GIP (gastric inhibitory polypeptide) and somatostatin. Rix et al (2015) do an excellent job of explaining what it is and how it works, so there's probably no reason to go on about it here. It will suffice to say that it is stored in granules and released on demand though exocytosis which seems to be a calcium-mediated process (i.e. intracellular calcium is the main secondary messenger system). Henquin et al (2017) found that there is about 0.7-0.8 mg of this stuff stored in a normal adult pancreas, which is approximately the same as a single adult dose for the reversal of accidental hypoglycaemia.  Its roles extend beyond purely making sugar available - it also stimulates lipid and protein catabolism, suppresses appetite, and increases total body energy expenditure, mainly by means of binding to G_s-protein-coupled receptors and increasing intracellular cAMP. For the intensivist, this latter function is of greatest interest, as it can be exploited to increase the contractility of the myocardium, bypassing the effects of β-blocker and calcium channel blocker overdose.

Insulin secretion by pancreatic β-cells

This gets a lot of attention elsewhere, so there is no reason to say more than this:

• β-cells secrete, store and release insulin.  In total, Bonner-Weir (2004) reports that an adult human pancreas contains about 200 units of insulin stored in β-cell granules, and secretes about 0.25-1.5 units per hour under normal conditions.
• Insulin is a 51-amino-acid peptide hormone arranged into two crosslinked chains
• The release of insulin is stimulated by:
  • Glucose, directly sensed by β-cells
  • Neurotransmitters from the autonomic nervous system
  • Incretin hormones
• The release of insulin is inhibited by:
  • Somatostatin
  • Ghrelin
  • Leptin
  • Galanin
  • Adrenaline
• The mechanism of insulin release, broadly, is:
  • β-cells have an ATP-sensitive potassium channel on their surface
  • ATP production leads to closure of this channel
  • This leads to membrane depolarization
  • Depolarisation produces the opening of voltage-gated L-type calcium channels
  • Intracellular calcium is the secondary messenger that leads to the release of insulin

This must represent some sort of bare minimum, so that one might still be able to say that this chapter on the endocrine pancreas is complete, but without duplicating a lot of content from other sections.

Pancreatic polypeptide secretion by pancreatic γ-cells

Now that we are done with the popular  hormones like insulin and glucagon,  what the hell is pancreatic polypeptide? Well, reader, it is a 36-amino-acid polypeptide secreted mainly from the islets in the head of the pancreas, and its role can be broadly described as "regulatory", insofar as it tends to interfere with a lot of different processes involved in digestion and metabolism. Like the other pancreatic polypeptides, its plasma half life is extremely short - only about 6 minutes. That's probably as much detail as a CICM trainee ever needs to know about this, as there is no possible way it could ever become a gamechanging SAQ in the First Part exam. However if you must know more, the best starting point would have to be this well-referenced Pancreapedia article by J.A Williams (2014). 

In brief,

• The release of pancreatic polypeptide is stimulated by:
  • Gut lumen content, i.e. the characteristics of the ingested meal, and specifically:
    • Protein is the most stimulatory
    • Fat is less stimulatory
    • Glucose is minimally stimulatory
  • Vagal stimulation, likely related to the volume of the meal
• The release of pancreatic polypeptide is inhibited by:
  • Ghrelin
  • Somatostatin (but then, it inhibits everything)
  • Atropine (by its anti-vagal effect)
• The mechanism of its action, broadly, is:
  • Inhibition of pancreatic exocrine and endocrine secretion
  • Inhibition of gallbladder contraction
  • Decrease in small intestine motility, increase in colonic motility
  • Inhibition of appetite
  • Increased hepatic insulin sensitivity and increased glucose uptake into hepatocytes 

This peptide is released mainly during the intestinal phase of digestion following a meal, particularly a large protein-rich meal, and its level remains elevated in the circulation for some hours. It has some subtle effects on the endocrine and exocrine pancreas (mainly inhibitory- it slows secretion), but its main effects seem to be on the regulation of satiety. Children with Prader-Willi syndrome have chronically low levels of pancreatic polypeptide, and are famously ravenous. In contrast, mice who have a chronic overexpression of this polypeptide tend to have poor appetite, and end up markedly underweight.

Somatostatin secretion by pancreatic δ-cells

Somatostatin physiology is discussed to some extent in the chapter on octreotide, the pharmacologically interesting analogue of somatostatin, and in the chapter on the physiology of pancreatic hormones.  In short, it is actually two cyclic peptides (a 14- amino acid and a 28-amino acid variant) which mainly inhibits the secretion of other hormones. It inhibits secretion wherever secretion is happening, be it endocrine or exocrine, and the family of somatostatin receptors are scattered widely, with some in the CNS. The pancreas is not a unique site of somatostatin production, as it could also come from the duodenum or pylorus, and in fact under normal conditions it appears the pancreas is a rather minor player. There is a pretty constant stream of it being released (albeit at a very low concentration) because this substance only has a circulating half-life of around two minutes. 

As is usual for all things pancreas, the Pancreapedia entry for somatostatin was the main source of this information. To summarise this already brief review:

• The release of somatostatin is stimulated by:
  • Gut lumen content, i.e. the characteristics of the ingested meal, and specifically a meal with a high fat content (Ensinck et al, 1990)
  • Cholecystokinin release
  • Increased blood glucose
  • VIP (vasoactive intestinal polypeptide)
  • Vagal stimulation, by some indirect mechanism (as the release is delayed by some minutes following vagal stimulation); but it can also inhibit somatostatin release
• The release of somatostatin is inhibited by:
  • Starvation
  • Muscarinic cholinergic agents 
  • Adrenergic antagonists
• The mechanism of its action, broadly, is:
  • Inhibition of secretion of basically everything everywhere, including but not limited to:
    • Pancreatic and pituitary hormones
    • Exocrine and endocrine secretions of the pancreas
    • Intestinal secretions, including gastric
    • Gallbladder contraction
  • Inhibition of splanchnic blood flow
  • Inhibition of upper GI motility

Most of the interest in somatostatin and specifically the replication of its effects comes from oncology, where it is used to block the secretory function of various adenomas, and for more detail the interested reader is redirected to the somatostatin entry from their no doubt dog-eared copy of the Handbook of Biologically Active Peptides (Kastin et al, 2013; page 548). For the intensivist, somatostatin (but more specifically its longer acting pharmacological analogue octreotide) is mainly interesting because they can be used to decrease splanchnic blood flow (reducing the rate of bleeding from the GI tract) and to reduce the volume of intestinal secretions (thereby allowing fistulae to close, or high-output stomas to calm). 

Ghrelin secretion by pancreatic ε-cells

Ghrelin, for the intensivist, falls into the realm of things we should be vaguely aware of, but only until something more clinically important comes along to displace the short-term memory from our brain.  It is a 28-amino-acid peptide, typically described as an orexigen. Rolling that term on one's tongue will soon produce the understanding that it is the antonym of anorexigen, i.e. a substance which produces weight gain by the stimulation of appetite. In fact the very word "ghrelin" apparently comes from the word root "ghre", meaning "to grow" in Proto-Indo-European language (Kojima & Kangawa, 2005) As the manipulation of appetite does not typically fall within the territory of critical care, the CICM trainee would be forgiven for ignoring this hormone completely, and learning more about the actions of adrenaline or something else more relevant to their craft. 

Still, its produced by the ε-cells of the pancreatic islets, and something should be mentioned here, mainly for completeness. Its release occurs in response to a diverse range of stimuli (which is another way of saying we do not yet have a clear idea of what stimulates its release). It appears to be released before each meal, and drops to its minimum levels approximately an hour after eating.  It is produced mainly in the stomach, with the pancreas contributing relatively little. Its effects include the release of growth hormone from the pituitary, increased secretion of stomach acid, increased gastrointestinal motility, and increased appetite; though it must be noted that knockout mice lacking ghrelin receptors appear totally normal.