RCC Hot Topics 2nd series
Managing TKI induced gastrointestinal toxicity in renal cell carcinoma (RCC)
Dr Ricky Frazer – Velindre Cancer Centre
Dr Sreedhar Subramanian - Cambridge University Hospitals
RF: Welcome to this discussion video. My name is Ricky Frazer, I’m a consultant at Velindre and I’m joined today by a great friend and colleague, Sree, who will introduce himself shortly.
What we want to do is think about how we manage TKI-induced GI toxicity in renal cell carcinoma.
And so I’m sure many of the audience will be familiar that TKIs have had a central role in treating kidney cancer. We’ve been using TKIs since 2006 but we know there are challenges. So one of those challenges are the toxicities that can happen and often these are adverse events that go on in the background and can detract significantly from quality of life and patient wellbeing. Over the years I’ve become a little bit more familiar with managing the different toxicities that we see but actually diarrhoea remains one of the main side effects that we see. So today I’m delighted to be joined by a gastroenterologist to talk about how to approach these.
As a reminder, when we’re discussing a case, this is how we grade it using the CTCAE and this is just looking at the different TKIs and in terms of the grade of toxicity that we tend to see. I think the take-home message is that with most of the TKIs diarrhoea is present to a lesser or more degree and so it’s a toxicity that we need to be really comfortable at investigating and then managing.
Again, this slide is really just showing that certain TKIs in combination, and we know that increasingly we’re using combinations in the frontline space, also have toxicity in the form of diarrhoea. But one of the challenges can be differentiating the cause of diarrhoea between the TKI or the checkpoint inhibitor.
So, Sree, I’d really value your thoughts on a case, if that’s possible. So this case is using pembrolizumab and lenvatinib but, of course, there are a number of IO/TKI combinations available in the frontline space and we know that diarrhoea is one of the toxicities that we see with all of these IO/TKIs.
So I had a patient back in 2014, had a nephrectomy for a clear cell renal carcinoma, all went to plan but then just under ten years later they developed multiple lung and liver metastases. I commenced them on a combination of pembrolizumab and lenvatinib. So Sree I really struggle with this if I’m honest, when I’m starting a patient on combination IO/TKI, are there any baseline stool testing I should be doing and again as an oncologist, maybe history taking’s not necessarily my strength. Is there anything that I should be asking the patient about in terms of their predisposing risk factors to developing diarrhoea.
SS: Yes, thank you Ricky. Just for introduction, my name is Sree Subramanian, I’m a gastroenterologist based at Cambridge. Even for gastroenterologists this is a relatively new field, we are just seeing increasing numbers of patients with diarrhoea, either due to TKIs or due to immunotherapy. So it’s been a steep learning curve for gastroenterologists all around. I would say that a little bit of attention to some history would be helpful, firstly to highlight the fact that there is a huge variation in bowel frequency even among a standard adult population. So some people open their bowels once a day, some people open them once every three days or some people go up to three times a day. So establishing the baseline stool frequency and seeing whether it’s loose or not can be an important thing.
In addition to that, as you highlight, some patients may have some relevant surgical history that could alter the bowel frequency. So, for example, things that could potentially affect stool frequency are things even relatively something innocuous like a cholecystectomy. If people have had their gallbladder taken out or if they have had a small bowel resection or a colonic resection, for example, it can increase the flow of bile through the gastrointestinal tract and cause bile salt diarrhoea. So in those instances patients can open their bowels multiple times a day and sometimes at nighttime. So if you know that a patient has already got a problem or if the patient has a history of a surgical resection, that could be quite important, particularly when the patient develops diarrhoea later on. So these are important things to establish, I would say. But other than that, I don’t think we need to do any stool testing as such, but just establishing some basic facts like stool frequency and surgical history is good practice.
RF: I think they’re really important learning points and certainly things I will take into the future. So, Sree, maybe if we can get back into this case. So this patient did develop diarrhoea shortly after starting the combination treatment. Interestingly, and I’m not sure I understood the significance of it, it was worse after eating. So these bowel symptoms, despite some changes, were ongoing 4-5 times a day. Their baseline, again, I obviously didn’t do a good enough history but their baseline was only once a day. Worse after meals, no blood, no significant urgency and did give some loperamide which helped with some of those symptoms. When I’m giving loperamide I’m thinking about the TKI, really. Clearly loperamide wouldn’t be the treatment of choice for an IO-induced toxicity. So I’m interested really just to think about at what point should I be picking up the phone and speaking to you? Is that management reasonable to do without a gastroenterologist or should I have reached out earlier?
SS: Obviously I’d never criticise your management, Ricky, so it’s expertly done. I would say the first-line management, it’s completely reasonable to use some antisecretory treatment such as loperamide. There are a few things in the history, again, which could help point towards a more severe problem with diarrhoea, for example if the patient has blood in their stools you are thinking that it may be inflammatory and it may be more likely to be due to the IO rather than the TKI. Likewise, if the patient has nocturnal stools, so for example if they are waking up multiple times during the night, then you know that, firstly, it’s more problematic from a quality of life perspective and, secondly, that there is probably more likely to be an organic process driving the diarrhoea. So these are two things to watch out for.
Clearly if the patient hasn’t settled down despite a reasonable dose of loperamide, that’s a good time to contact the gastroenterologist. But before that maybe some basic stool sample tests like infection, C. diff., some of these patients might have had a course of antibiotics.
Interestingly, this is a recent example of a patient that I’ve seen where a patient had a cytoreductive radical nephrectomy and as part of that she lost most of her pancreas as well and she had some small bowel resection due to an accidental enterotomy. So you can straight away see that that patient might clearly have an alternative explanation for the diarrhoea, such as pancreatic insufficiency. So even though we talked about baseline surgical history, I think it’s important to just keep on top of that and make sure that the patient hasn’t had any other interventions that might drive the diarrhoea.
RF: Interesting. I think as an oncologist I’m often just assuming it must be the TKI, I think that’s how I’m programmed so it’s interesting that you’ve mentioned those other causes. So maybe if I can hand over to you, Sree, because I think there’s some really interesting data out there now which probably pushes back on me just assuming the TKIs are the cause.
SS: Yes, absolutely, thank you. Much of the credit for the next few slides is due to Kate Fife, who is an oncologist based at Cambridge and she in conjunction with a Fellow of hers looked at a series of patients who had been referred to a local gastroenterologist for really troublesome diarrhoea. So I think she referred approximately 30 patients and then quite a lot of them had some simple work-up, nothing fancy, no colonoscopy or no invasive investigations. We found that a lot of patients actually had secondary causes. I already mentioned bile salt diarrhoea due to disordered bile acid homeostasis and some patients had pancreatic insufficiency. Pancreatic insufficiency, for example if a patient has got a history of diabetes, you’re much more likely to think about that. A couple of patients had small bowel bacterial overgrowth. We’ll cover these in some more detail in the next few slides. So, if the patient has intractable diarrhoea it is definitely worth considering secondary causes.
RF: Interesting, and they’re not causes that would jump to my mind, I have to be honest, so I’m looking forward to being educated on them. So if we come back to this case, and this is just a reminder of the symptoms the patient presented with, based on what you’ve just told me, what investigations should I be requesting and why? It’s easy now I’m speaking to you but life is busy and clinics are busy, so out of those tests you’ve just discussed, and I know we’ll get into them in more detail, can I request them? Do I need you to request them?
SS: No, of course. It’s always helpful if there are some baseline tests that have been done by the time they come to see a gastroenterologist, partly because they could be waiting a couple of weeks at least, or longer, before they get to see one, and there are some relatively simple low-hanging fruit type of investigations.
For example, that will include stool culture, C. diff., stool for calprotectin I think is a really important test, especially in the context of somebody who is on combo TKI and immunotherapy because this diarrhoea could be inflammatory. Even though there is no blood in the stool it could still be inflammatory and therefore calprotectin is a protein that’s found within neutrophils, it’s a calcium and zinc binding protein, so wherever there are neutrophils the calprotectin is going to be elevated. So you can measure calprotectin in literally any bodily fluid – it can be in the stool, it can be in the serum. So that can be really helpful. So if the calprotectin is normal then you’re less inclined to think that it’s the IO and more inclined to think that it’s a non-inflammatory diarrhoea. And likewise a faecal elastase is also a good, simple test to do. That measures pancreatic exocrine function. So these are simple, non-invasive tests that can be done by an oncologist ahead of referring to a gastroenterologist.
RF: Okay. So, in this case, I have to be honest, I did reach out to a friend and got some help in terms of what to investigate. This is what came back. So then the next challenge for me, Sree, is knowing what to do with the results when I get them. So I’d really appreciate just adding some context of what this means.
SS: Of course, yes. So basically the magnesium being low is not particularly surprising, that can often be driven by diarrhoea itself. The faecal elastase is a bit borderline, so in most laboratories a faecal elastase of more than 500 is completely normal, it excludes pancreatic exocrine insufficiency. If the faecal elastase is between 200 and 400 that’s a little bit borderline. One of the caveats with faecal elastase is that if the stool is loose, say for example if you have diarrhoea from any other cause, the faecal elastase can be falsely low. So I always feel that it’s worth retesting the faecal elastase at the time when the stools are a bit more formed. Now, that can be difficult but it is good practice to re-check it.
We spoke about the calprotectin, here the calprotectin is reassuringly normal and what we would say is it’s probably non-inflammatory diarrhoea. You can be fairly confident about this with a good degree of sensitivity. The SeHCAT scan is a scan which is designed to look for bile salt malabsorption, so it’s a radiolabelled bile capsule that you swallow on day 1 and then you sit in front of a gamma camera that takes a picture and you come back a week later and have the same x-ray done and then if the retention is less than 10% it indicates that the bile salt recycling is not working properly and therefore indicates that bile salt malabsorption may be a problem. I think we have got more on that in the next few slides. So here we can be reasonably sure that the patient may have bile salt diarrhoea.
RF: So I think you’ve highlighted, I have to be honest, a couple of things that I’m not that familiar with. So it would be great if you could maybe just walk us back through those because certainly hydrogen breath test is something I think I learnt as a medical student and have thought nothing else about ever since.
SS: Sure. So here are the few things that we spoke about which are fairly simple, non-invasive tests. We have spoken quite a lot about the faecal elastase. Yes, so taking it in turn, bile salt malabsorption, this is something firmly back from medical school days, I would imagine, for most people, probably even for gastroenterologists. Essentially you’ve heard of the concept of enterohepatic recirculation where the bile is a slushy yellow liquid that’s made in the liver and is stored in the gallbladder. In response to a meal the gallbladder squeezes out the bile which then mixes with the fatty food and helps digest the food. Together the bile with the food travels along the length of the small bowel and when it gets to the terminal ileum the bile is reabsorbed through the enterocytes back into the liver.
When it is reabsorbed there is a negative feedback loop which is mediated through FGF, fibroblast growth factor, which is part of the broader TKI family of receptors. Through this negative feedback loop it stops the liver from making more bile. But actually, for example, if the FGF is inhibited for whatever reason then you can see that actually that negative feedback loop is not working, more bile is being produced and the terminal ileum is overwhelmed, it can’t just simply reabsorb the amount of bile that’s slushing through.
It can happen for a variety of reasons – if the patient has had their small bowel taken out, for example, the terminal ileum taken out, particularly if a patient has a history of Crohn’s disease and they’ve had surgery.
It can be fairly simply fixed with bile salt binding medications such as cholestyramine or colesevelam which can turn the bile into a fairly innocuous substance. The problem with bile is that it actually increases the amount of colonic water secretion and it causes this typical postprandial diarrhoea. So most of the bile comes down after eating, so you can see that patients are going to rush to the toilet after they’ve eaten something. So it’s quite an important thing to exclude.
RF: Because that was exactly the story the patient gave which I’m not sure I really accentuated when I gave you the history. But that was exactly what the patient was saying to me. – Maybe just, I know that we’ve talked about this already Sree but maybe just summarise this for me again, that would be great.
SS: Yeah just a quick one minute summary. The pancreas has two main functions, exocrine and endocrine. The exocrine digestive function can be checked simply by measuring the faecal elastase which is a simple test. You may remember back when we were in medical school we had to do a three-day faecal fat collection. I’ve certainly done that as a junior house officer, it’s not anyone’s idea of fun, but that’s been replaced thankfully with faecal elastance. The caveat that I mentioned, if the stool is loose please do recheck it. But if the patient’s had history of chronic pancreatitis or previous pancreatic surgery this is definitely a contender for causing diarrhoea, so worth checking it and again you can replace it quite simply with pancreatic enzyme replacement therapy.
RF: Brilliant.
SS: Small bowel bacterial overgrowth is a bit more of a niche thing. We spoke about the hydrogen breath test. Now, the gradient of bacteria increases along the length of the gastrointestinal tract. There are a lot of bacteria in the mouth, the stomach is relatively sterile because of the stomach acid but then the amount of bacteria gradually increases along the length of the small bowel, reaching a peak in the colon. If there is more bacteria in the small bowel potentially it could interfere with the food absorption and cause diarrhoea. Now, the problem is the test that we use for small bowel bacterial overgrowth is an indirect test. So you give the patients glucose and then that’s obviously going to be metabolised by gut bacteria and it releases hydrogen as a byproduct of glucose metabolism. If that hydrogen is being released early on, that means that glucose is getting metabolised higher up in the gastrointestinal tract which indicates a patient may have too many bacteria in the small bowel.
Now, it’s much more likely that a patient may have small bowel bacterial overgrowth if they have had a history of intestinal surgeries where there has been replumbing and therefore… For example, if the patient has had a right hemicolectomy, the ileocecal valve has been taken off, there is direct communication between the small and the large bowel, those are situations where you can have a high index of suspicion patient. But otherwise it can be a slightly contentious diagnosis so even if you do make it, even if you do have a positive hydrogen breath test, it’s worth critically thinking whether this patient truly has small bowel bacterial overgrowth.
RF: So, based on the fact that my patient, on reflection, had a really clear history of bile acid malabsorption, how do we manage it?
SS: So the treatment for bile salt malabsorption is fairly straightforward. You can give bile salt binders which actually suck the bile, bind to the bile and turn it into a fairly innocuous substance. As I mentioned, much of the bile comes down after you have eaten something so it’s important that the patient takes the medications at least half an hour before they eat a main meal. It’s very important as well to avoid taking any other medications for an hour afterwards because although these medications are meant to bind bile, they are quite non-selective anion binding regimes so they can bind to other medications and make them inert.
RF: Really useful. So we’ve talked a little bit about the management of the bile acid problem. Maybe, which will be very straightforward for you, I know, this stuff, Sree, but some guidance on diarrhoea management in general?
SS: Yes, sure. I think once you’ve excluded inflammatory diarrhoea and once you’ve excluded infection, straightforward antisecretory treatments like loperamide are a perfectly reasonable thing to do and you could use 2mg after every stool, for example, just to slow down the bowel. Dietary modification is a bit more hit and miss, particularly if the patient has got ongoing diarrhoea. It’s definitely worth excluding the usual kinds of suspects such as, for example, cutting down the amount of dairy and lactose, animal milk for example. These are things that can be tried. There are more esoteric things like the FODMAP diet which is cutting out fructans and oligosaccharides but that probably would need the help of a dietician.
Then, of course, it’s down to the oncologist to review the oncology treatment, see if there is anything that could be done. That’s more your territory rather than mine. All the while also bearing in mind at the back of our minds that the patient is on combo treatment, so worth making sure that there is not an IO component to this.
RF: So we focussed primarily on TKIs today but in terms of the immunotherapy component we know that immunotherapy can cause diarrhoea. Is there any difference in the way that it presents or, indeed, how we might investigate that?
SS: Yes. History-wise I think it’s a bit more tricky and it’s difficult to be certain. Sometimes there can be no major distinguishing factors between the two, to be honest with you. Blood in the stool is definitely a sign, especially a nuance that blood in the stool is definitely a sign that there is an inflammatory process or a colitis that is going on. Again, nocturnal stools can be an indicator that there is something organic, something inflammatory quite possibly. But other than that there are not many features. If the patient just has watery diarrhoea you’re shooting in the dark a little bit. I think here is where the role of calprotectin cannot be overstated – definitely worth checking that just to make sure. Calprotectin is not a perfect test, it can be elevated by a number of different things, including non-steroidals, proton pump inhibitors and so on and so forth. So I think we need to be cautious about interpreting it but, nonetheless, a negative result can be helpful in this context.
RF: In terms of managing and investigating, but primarily managing, the gastrotoxicities that we see with IO/TKIs, are there any resources available to support clinicians?
SS: I’m sure Oncology Society guidelines are just as helpful and informative but at the end of the day you can’t beat having a local, friendly connection with a gastroenterologist and a liaison, just to go through the different cases or particularly if you have a problematic case that’s not settled down, that kind of email connection or a local connection with a gastroenterologist is really helpful.
RF: Great. So this patient actually continued then on combination after we identified the bile acid malabsorption and managed it and has had no further episodes of unmanageable diarrhoea. A real learning for me not just to assume it’s TKI related but to think about some of those other causes. So if I maybe just summarise, what I think we’ve talked about today is thinking about TKI-related diarrhoea but actually thinking about some of those other causes that are manageable and actually may allow our patients to stay on treatment. Ultimately when I see patients in clinic that’s what they want – they want to stay on their treatment. We do need to think about how to manage that diarrhoea and you beautifully articulated some of the other ways that we can manage it. And I think we’ve acknowledged the fact that there can be challenges in differentiating between TKI or checkpoint inhibitor induced diarrhoea, but actually that there is some really robust guidance available to us. So, Sree, I just want to say a massive thank you for walking me through that and holding my hand and I look forward to discussing lots more cases along the way.
