A source for healthcare professionals to access the latest data and information on the diagnosis, treatment and management of patients with gut related disorders

Video series

A series of video presentations exploring information relevant to gastroenterologists, including summaries from key conferences.

Pathophysiology of bile acid diarrhoea
Professor Julian R.F. Walters

Professor Julian Walters discusses the pathophysiology of bile acid diarrhoea with regards to ileal disease and ileal resection; he also discusses primary idiopathic bile acid malabsorption diarrhoea.

Read Transcript >

So what is the pathophysiology? In ileal disease, malabsorption of bile acids occurs in ileal disease or ileal resection, so the bile acids are not being recycled. But in primary idiopathic bile acid malabsorption diarrhoea, people have not been able to show a defect in ileal bile acid absorption. No defect in bile acid transporters, there is a larger bile acid pool size and an increased bile acid synthesis, and I’ll show you the evidence for these things now.

The important thing is that hepatic bile acid synthesis is under negative feedback control by the ileal hormone, Fibroblast Growth Factor 19 (FGF19). This is a hormone that I hadn’t heard about until 2005, but then there was a eureka moment when I realised that this might be responsible for causing the pathology of primary bile acid diarrhoea (BAD).

So let me show you there’s no bile acid absorption defect in Idiopathic bile acid malabsorption, primary BAD. This paper is from Van Tilbergs group in 1991, so quite a long time ago, taking patients here basically finding no defect in uptake into brush border membrane vesicles.

A second paper from Van Tilburg looked at bile acid pool size and faecal bile acid loss, so here’s our patients with primary BAD, here. Their faecal bile acid loss is about two and a half times the controls. Surprisingly, ther bile acid pool size is about twice as big in the patients as in the controls. And you can see the SeHCAT retention, which here is expressed as a half-life in days, so they have a shorter half-life and this is the paper that shows the bile acid pool size can be larger in these patients and you can imagine the difficulty they had in publishing this. This is in the Scandinavian Journal of Gastroenterology Supplement, because people said if you have malabsorption you should have a smaller pool size. They have found a bigger pool size.

So this fits in with FGF19 and FGF15. FG15 is in mice and FGF19 in humans, and it’s an enterohepatic signal that regulates bile acid homeostasis. So you can see it’s found here in the ileum and it goes up with a certain agonist; GW4064. Effects are dependent on the bile acid agonist dependant gene in the ileum, much more than all these other genes. And in these animals here that have had FGF15 knocked out, they had an increased amount of faecal bile acids.

So FGF15 in mice, FGF19 in humans; made in the ileum. Bile acids bind to this receptor fXR. Farmacoide X receptor (FXR), and produce FGF19. So it’s a transcriptionally active event. FGF19 goes into the portal venous blood, goes through the hepatocyte, where it binds to the receptor FGFR4, and the co-receptor beta-Klotho, and will inhibit bile acid synthesis by CYP7A1.

So could defective FGF signalling be the cause of primary BAD? We had this paper in 2009 which shows a new mechanism for BAD: “Defective Feedback Inhibition of Bile Acid Biosynthesis” and there was an editorial by Alan Hoffman, David Manglestof and Steven Clever in the same group. Chronic diarrhoea due to excessive bile acid synthesis and not defective ileal transport. So in our patents, we’ve got 17 patients and 19 healthy controls. Most of our patients with diarrhoea have had a SeHCAT test and you can see the C4, the median C4 is higher in our patients than in our controls. So this had been shown before. There is a wide scatter here and this shows they are producing more bile acids. We looked at the FGF19, and the median FGF19 is about half in our patients compared with the controls. Quite a bit of scatter, both in the controls and in our patients but this is a significant median difference here.

So what do we think is happening is here? In a normal situation bile is coming down to the ileum, the bile acids are being absorbed, they are binding to FXR and producing FGF19, that goes via the portal blood stream to the liver, where it inhibits CYP7A1. So cholesterol , which normally goes to C4, goes to bile acids and feeds in here. So the synthesis here is equal to the amount that gets in to the colon. In our patients with BAD, although the absorption is fine here, it may even be more, there is failure to produce FGF19, so there is less feedback inhibition, so there is more synthesis of C4, and then bile acids, so you’ve got more bile acids coming down here and then more bile acids entering the colon and this then leads to secretory diarrhoea.

So we did a prospective study of our patients who came in, Pattni S; Walters JRF. Br Med Bull. 2009, and we got chronic diarrhoea controls, so SeHCAT negative patients (patients with SeHCATs more than 15) and our primary BAD group with SeHCATs less than 15, and again you see the scatter here, but again there is a significant difference in the median here in our patients with primary, and our patients with secondary BAD compared with the chronic diarrhoea controls.

When we looked at the primary BAD group compared with the controls, you can see a range of features down here. The female predominance is less in primary BAD than it is in the controls. These people are having a median of 6 bowel movements per day and up to one at night. The duration of the diarrhoea is many months, so fitting in with David Sanders work, the median duration is 24 months. And they’ve got incontinence, urgency, abdominal pain and bloating in various proportions. Low FGF19 will also predict the patients who were more likely to respond to bile acid sequestrants, cholestyramine or colesevelam. So patients with this cut off here were more likely to be responders.

So lets look at where we are now. In health, normally, the ileum absorbs bile acids and makes FGF19. These are patients with idiopathic BAM, or as I prefer, primary BAD. So the absorption is going on but FGF19 is not being made, so there is excessive bile acid synthesis. In patients with ileal disfunction or resection, there is excessive bile acid synthesis, there is failure to absorb bile acids and presumably failure to produce FGF19.

So I won’t say very much about Crohn’s disease because we will be hearing about Crohn’s disease later on, but we looked at patients with documented lengths of ileal resection; those who have had a longer resection have a lower fasting FGF19 level. So one of the defects that occurs in Crohn’s, is the malabsorption of bile acids but perhaps as important is the failure to produce the FGF19. So maybe actually in Crohn’s we’re seeing the same situation that we see in the primary patients, failure to produce enough FGF19.

Jonathan Nolan was my research fellow and we looked at a range of different patients here; so we’ve already seen controls and we’ve seen idiopathic diarrhoea patients. This is non-resected Crohn’s patients without diarrhoea, so they are significantly lower. If they’ve got diarrhoea, it’s lower still. If they’ve had an ileal resection it’s low and if they’ve had an ileal resection with diarrhoea, it’s low.

So the lowest levels occur in patients with ileal resection and diarrhoea and it correlates with the activity of the disease and the Harvey Bradshaw Index.

Briefly we have looked at patients with Crohn’s disease who have had treatment; those who’ve got colonic disease who have received treatment with either steroids or anti TNF’s had no change in their FGF19, but if they have ileal disease, and they have treatment, then the FGF19 will go up. So inflammation in the ileum is depressing the FGF19 in patients with Crohn’s Ileitis.

So this is a figure that comes from Michael Camilleri looking at the various genes involved in the liver and in the ileum, and we have looked at some of these variants in this bile acid FXR, FGF19, FGF receptor 4, beta-Klotho pathway.

So in our patients we’ve got our 70 or so primary BAD patients, and our 80 or so diarrhoea controls, and we’ve looked at the range of genes, including FXR, FGF19, the receptor, Beta-Klotho, ASBT, which is the aprical sodium dependant bile acid transporter, and OST-Alpha, which is the basolateral membrane transporter.

Now Michael Camilleri has reported significant differences in variance with FGFR4 and beta-Klothro, but we do not find in our patients any significant associations with these numbers. So this is not going to be a major phenomena; we can’t rule out that in a sub group of a few patients that there are genotype differences that may affect the responsiveness of the liver through the FGFR4 and beta-Klotho pathway

We’ve looked at some of the expression of these genes and we have related these to SeHCAT. Patients who have had colonoscopy have had ileal transcripts taken, and you can see we’ve looked at ASBT RNA and there is a correlation here of SeHCAT values and ASBT. Similarly, we have looked at SeHCAT and FGF19 and again there is a correlation here. And there is a correlation between FGF19 and ASBT, but there is quite a range of different ASBT RNA levels for low FGF19 levels. Johnston et al. is a paper that is in press in the American Journal of Gastroenterology.

Now we’ve also taken ileal biopsies, and we’ve done explants with a 6 hour incubation with different bile acids, and then done RT-PCR to look at expression of various genes, and you can see the stimulation of chenodeoxycholic acid or the glychochenodeoxycholic acid, FGF19 in this model is by far the greatest stimulation. Over 300 here compared with some of these other genes which are known to be bile acid FXR response genes of only two or three. We looked at the time course, we looked at the concentration dependence and we were also able to show that the protein in the tissue culture fluid goes up. So when we looked at our patients again, we’ve got about 16 patients who had SeHCAT tests and have had stimulation. There is a correlation here between the SeHCAT values and the stimulation of FGF19 by chenodeoxycholic acid, and you can see there are some patients here with very low SeHCAT values who have very little stimulation.

Again, you can see FGF19 is by far the most potent stimulator gene, other ones are only two- or three-fold. There is good correlation of FGF19 with SeHCAT, and there is this correlation also for the amount of IBABP, the intracellular binding protein with SeHCAT, but not with these other genes.

Finally, I want to talk a little bit about treatment for BAD. So you will hear about bile acid sequestrants again later, but would FXR agonists be a treatment for BAD? They would stimulate FGF19, and that would then lead to inhibition of hepatic bile acid synthesis, and this could then reduce the colonic secretion and the symptoms. So we’ve looked at obeticholic acid, which a company called Intercept is promoting for liver disease, you can see that obeticholic acid is a more potent stimulus than chenodeoxycholate. This is a log scale here. So 1uM obeticholic acid is similar to 50uM chenodeoxycholate, and 20uM obeticholic acid gives many fold greater stimulus than chenodeoxycholic acid.

We had this paper in Ailmentary Pharmacology & Therapeutics last year, “The response of patients with bile acid diarrhoea to the Farmesoid X receptor against obeticholic acid”. Briefly, patients took this for two weeks and you can see their fasting FGF19 went up from about half normal to a normal value. So this was a dosage that would lead to most patients having a restoration of FGF19. The C4 would fall from raised levels to a lower level, and the total bile acids produced in the area, and the curve after meals, would also fall. This also produced a clinical response - so there is still frequency - this is a median stool frequency per week, so about 21 stools per week went down to 14, just with a two-week treatment. The stool type improved from a median, or worse than, type 5, to better than type 5; And we worked out an index that combined these, and the loperamide dosage, and the index improved, and in fact the index improved in every patient that we treated. These were all SeHCAT positive patients.

So to sum up, let’s look at the pathophysiology of primary BAD as an endocrine disorder. So the main symptom is diarrhoea, the direct cause is excess faecal bile acid. The pathophysiological problem is unregulated BAM, and we have a hormone that regulates this metabolism, FGF19. We have a defect sometimes on feeding; impaired production on feeding, and there is the possibility of impaired receptor function. So compare that with the classical endocrine disorder, diabetes, where the symptoms are often produced by polyurea due to excess urinary glucose, with unregulated glucose metabolism. We have a hormone, insulin. We have impaired production in type 1 diabetes and we have impaired receptors in type 2 diabetes. So I think much of primary BAD can be seen as an endocrine disorder.

If you think about the diagnostic strategies, we can measure hormones, so we can measure FGF19, but a bit like insulin it goes up and down, and it may not be as accurate as some measurements. We can look at the regulated product. So we can look at bile acids or C4, which is a bit like measuring glucose in diabetes. The pathological problem is excess bile acid secretion or faecal bile acids vary cyclically after meals, as they are dependent on synthesis, absorption and hormonal action. So we’ve got hyperglycaemia and glycosuria as similar, all go up and down depending on meals.

We have tests though that integrate function over multiple cycles of meals of production, reabsorption and secretion. So we’ve got SeHCAT which is a 7 day test, and so will integrate many meals worth, this can be thought of as being similar to HbA₁꜀ in diabetes where we are looking at something that changes over many days.

And then if you think about the therapeutic strategies where we can reduce the amount of the hormonally regulated product, so a low-fat diet in primary BAD would be a bit like a low carbohydrate diet in diabetes. We could treat the effects of excess hormonally regulated product, so bile acids sequestrants would do that, and that’s a bit like the fluid and electrolyte replacement for diabetes. We could stimulate hormone production, so sulfonylureas will do that in diabetes, and FXR agonists like obeticholic acids would do that in primary BAD.

I haven’t got anything to look at the down-stream effects of the action of the hormones, so like metformin in diabetes, and at the moment we haven’t yet got to the stage where we can replace the hormone as we do with insulin, but there is a company that has made a synthetic FGF19 which is going into early trials.

So the mechanisms affecting BAD are complex and I had this review in Nature Gastroenterology last year. Lots of different pathophysiological processes at lots of different sites. We’ve mentioned some of these factors down here but we haven’t mentioned many of them. There’s other factors, microRNAs, Inflammatory cytokines, a gene called diet1, the microbiome is important. The overall response though will also be critical. So the amount of visceral hypersensitivity and indeed the psychological response will be important. Do patients see passing four wet stools a day as being a problem or is that just what they’ve always done and something they can tolerate.

Job number: JB57410GBf Date of Preparation: June 2019

Let's talk about irritable bowel syndrome (IBS)
Professor David S. Sanders 

Professor Sanders highlights that many patients presenting with irritable bowel syndrome (IBS) have underlying diseases. He explores why  patients do not always benefit from having treatments that are well recognised for IBS and notes that some patients may in fact have another diagnosis, disease or disorder.

Read More


Mechanism of bile acid diarrhoea
Professor Julian R.F. Walters

Professor Walters explores the classification of different causes of bile acid diarrhoea (BAD). Professor Walters describes the normal entrohepatic circulation of bile salts and what occurs when the system does not function correctly. 

Read More


How common is bile acid diarrhoea (BAD)?
Professor David S. Sanders 

Professor Sanders argues that many patients presenting with irritable bowel syndrome (IBS) have underlying diseases. He explores a series of studies that suggest approximately 25% of patients presenting with IBS symptoms have BAD as the underlying cause. 

Read More

Bile acid diarrhoea: who should I test 
Professor David S. Sanders

Professor Sanders discusses which clinical scenarios might warrant investigation for bile acid diarrhoea (BAD) and how this could improve treatment for these patients.

Read More


Irritable bowel syndrome (IBS) is a disorder of exclusion: An extensive diagnostic work-up is necessary
Professor David S. Sanders 

Professor Sanders highlights that many patients presenting with irritable bowel syndrome (IBS) have underlying diseases. He explores why patients do not always benefit from having treatments that are well recognised for IBS and notes that some patients may in fact have another diagnosis, disease or disorder.

Read More


Are you a healthcare professional?

This educational website is developed by GE Healthcare for educational purposes. The information contained within this website is intended for healthcare professionals. By entering this website you are confirming that you are a healthcare professional.

Enter the website

This link opens another website that is not under the review or control of GE Healthcare Ltd and, as such, the company does not endorse the content, its accuracy, or any practices or standards contained within it.