This is the blog post version of a talk given by Andrew Weatherall (@AndyDW_) at the Paediatric Anaesthesia Congress of South Africa, 2015 held in jacaranda-bedazzled Johannesburg. There were no financial disclosures or conflicts of interest to put out there.
When I was first given this topic for it read like this:
“Anaesthesia for the Patient with Hepaticopancreaticobiliary Disease having Non-Hepaticopancreaticobiliary Surgery”. Phew. Bit of a doozy.
So I changed the topic but I think the aim is still pretty much in the ballpark.
I have three jobs. The first is to remind everyone that rather than just being an unappetising bit of offal in the upper part of the abdomen, the liver really is an organ that’s vital to everything.
The second is to cover a little bit about why anaesthetists should care if it’s got a bit of disease. After all there’s so much reserve in the liver, do we really have to care?
The third is to get into the nuts and bolts – what are the effects of liver disease we have to be aware of and how do they guide our anaesthetic?
Consider a Story
The whole way through I think this will be easier if you can picture a patient. A patient in front of you that needs your help. So let’s imagine an 8 year old. She has end-stage liver disease due to biliary atresia. On your assessment she has evidence of portal hypertension with a big spleen and hypersplenism contributing to a Hb of 89 g/dL, a white cell count of 2.8 and a platelet count of 84 x 10(9)/L. Blood testing suggests mild synthetic dysfunction with the INR at 1.5 and albumin 29. Your clinical acumen is pretty advanced and you manage to pick up mild ascites.
Of course patients with liver disease get sick and she has appendicitis. The surgeons would like to operate and she needs anaesthetic care. That care will be a lot easier if we really know the liver.
Renewing an Acquaintance
The liver used to be a big deal. The ancient Greeks thought the liver was so important that one of the worst torments of the underworld concocted was to have your liver eaten by a giant eagle every day for the rest of forever. Poor Prometheus. Never mess with the king of the gods.
(As a sidenote, that suggests the Greeks came up with the idea of liver regeneration a long, long time ago.)
Since then people seem to have become fascinated with other organs like the brain and the heart. They’re fine I guess, but there’s a reason the liver’s one of those organs we can’t replace with machinery. It’s just too good at too many things.
So let’s not just think about it as a giblet.
Let’s think about it as a voracious digesting machine, a bit like Cookie Monster, helping through the glorious production of bile and letting us digest and absorb all those fat-soluble goodies.
Or let’s remember that the liver is really the Lego master builder of the human body. It breaks big things down and builds things back up in all nutrient groups – carbohydrates, fats and proteins. It metabolises drugs into smaller chunks in phase I reactions then repackages for excretion in the phase II reactions.
So let’s give some credit to the liver. The monster master builder of the human body.
This metabolic skill set extends further though. Like the obsessive compulsive banded cleaner shrimp, confronted with toxins of all sorts, the liver tries to clean its environment. This is pertinent to the things we present it with as well as the later discussion on things that alter patient neurology.
So we have the liver – the monster super cleaning master builder of the body.
The liver also does hormones of course. It is involved in production of EPO, 25 hydroxycholecalciferol, steroid hormones and somatomedins to name a few. The liver monster super cleaner weightlifting master builder of the body.
There are different ways to seem burly though and the liver has a role in saying “you shall not pass” in the manner of a particularly prominent hobbit-befriending wizard. The liver has important immune functions as it participates in the reticuloendothelial system. The Kupffer cells on the sinusoids are pretty vial in antigen presentation. So he ever is actually the monster super clean weightlifting master builder wizard of the body.
Then after all that the liver also stores things. Things like iron, B12 and glucose. Oh, and blood. Glorious, glorious blood.
So there you have it – how about a little respect for the monster super cleaning weightlifting master building storage wizard of the human body?
Well like it says, so what? Well there’s a definite increased risk in the perioperative period for these patients. The thing is a lot of the literature goes over things like grading scales to assess risk. So you might read descriptions relying on the Child-Turcotte-Pugh score (which is really better for adults) where the scoring involves assessment of encephalopathy and ascites amidst a range of parameters. You can divide patients into classes A through C (with C the more severe disease) and recent studies suggest mortality rates of up to 12% depending on the underlying operation. Is a grading scale really useful to the anaesthetist conducting their own assessment though?
The Pediatric End-stage Liver Disease scale is really designed for those under 12 awaiting transplant. To calculate it requires inclusion of sex, age, growth items, albumin, bilirubin, INR, coagulation studies and some diagnostic items). For all that it’s still not strongly associated with outcomes.
Where does that leave the clinical anaesthetist? Something that’s easier to think about is the surgery itself. There’s at least some literature out there suggesting abdominal surgery mortality rates of up to 30%, cardiac surgery mortality rates between 16-31% and trauma laparotomy mortality of around 45%. Those are big numbers and enough to say the elevated risk is real.
The story for the anaesthetist with the 8 year old with appendicitis in front of them might be easier if we work through things we understand. So let’s work through body systems.
Bits of the Body
There’s really 3 key bits to consider with the lungs. One is the simple mechanical changes that can be induced. The other two are very particular conditions relating to the liver disease.
The mechanics are easy to grasp. Some of these patients have ascites. Ascites may also be associated with pleural effusions. Some have organomegaly. When this is the case you can end up with a reduction in lung volume with an associated drop in functional residual capacity. You end up with lung units operating in the zone where small airways close. Operating in that zone is a recipe for increased work of breathing, less effective muscle function to produce ventilation, and inefficiencies in ventilation and perfusion matching. The anaesthetist faced with this may want to consider closely the positioning of the patient to improve the mechanical advantage of respiratory muscles and improve the status of the lung volume.
An entirely separate condition is portopulmonary hypertension. This looks like idiopathic pulmonary arterial hypertension but wiht the prerequisite of splanchnic disease. While many of the anaesthetic approaches (control carbon dioxide, deliver oxygen, avoid sympathetic outflow etc) look the same one thing is different – some treatment options are off limits due to the portal hypertension. Add calcium channel blockers and you will cause dilation of the splanchnic bed.
Perhaps the most intriguing of the three is hepatopulmonary syndrome. This consists of a triad of portal hypertension, widened A-a gradient and small intrapulmonary vascular dilations which represent a shunt directly from the pulmonary arterial to pulmonary venous circulation.
The thing that is a little particular in this setting is that these patients have improved oxygenation when you lie them flat. Sit them up and their SpO2 reading will be in the low 90s. Lie them flat and it goes up 4% or more.
This makes more sense when you know that those dilated bits tend to collect in the lower zones, exactly the areas getting the most blood when you sit up. That’s a whole lot of shunt you remove when you lie flat. You can demonstrate this diagnosis with a delayed bubble test (introduce agitated fluid to a vein and watch for it to reach the left atrium too slowly for it to be another site of shunt) or nuclear medicine testing.
So for our 8 year old? She has evidence of ascites and organomegaly. She has neither of the more colourful conditions .Perhaps we should be ready for those mechanics.
I grew up being told that these patients where at great risk of bleeding because they were coagulopathic. The INR is high after all.
Well, that was wrong.
It’s now understand that while there are things not produced by the diseased liver that predispose you to clot as well. So for all the known issues of low coagulation factors, low numbers and function of platelets, dysfibrinogenaemia and more tPA, we need to acknowledge the other things not produced. We have to remember that there is more von Willebrand factor and factor VIII, less protein C and S and less antithrombin III (amongst other things).
Perhaps this plains why you might have met those patients with an INR that’s a bit up who just don’t bleed at surgery. For the clinical anaesthetist that means you really have to rely on clinical assessment, not just blood tests for test tube coagulability. These patients do bleed, but that is most often because of physical factors like their varices or portal hypertension.
This all applies to our 8 year old too – being prepared is important, reversing lab numbers might not be.
You might well recall the appearance of a patient with liver disease who is just in a bit of a fog. This is classically when they are acutely unwell. The liver, not quite producing it’s cleaner shrimp special skills, gets confronted with ammonia absorbed from the GIT (originally generated by the bacteria that reside there) and lets it all pass. It reaches the brain, crosses the barrier and ends up in astrocytes.
Of course those astrocytes don’t know what’s hit them, turn it into glutamine but the causes swelling of those rather important cells. Hence the picture of poor neurotransmission, oedema and eventual coma.
We’re familiar with that but a little aside – given two patients with chronic liver disease, kids with early onset of that disease tend to have a lower IQ than those with late onset disease (measured in a window 8-11 years later). Over half have some form of delay in neurological development. That’s probably worth factoring in when you try to figure out exactly what their current neurological status is compared to how they might usually look.
Our patient today? She is pretty vague on a bunch of simple things and seems a bit drowsy. Of course, she’s been up so she could just be tired, right?
The Cardiovascular Bit
The classic picture of these patients is a high output, low resistance circulation. Most things you read will say this might be due to more endogenous vasodilators going about or decreased clearance of those little chemicals (think NO, CO, endogenous cannabinoids etc). This is significant in practice because these patients won’t tolerate hypovolaemia well. Fluid management will be a key challenge.
Meanwhile the body also retains sodium and water and any portosystemic shunts increase the venous capacitance.
This describes your 8 year old with appendicitis too. Her heart rate is about 135/minute and her blood pressure is 90/35. Happily she doesn’t have one of the nasty liver conditions associated with direct myocardial impact (think haemochromatosis, Wilson’s disease or amyloidosis).
That hypovolaemia bit, that matters. Like in our 8 year old, who has had vomiting with her appendicitis. Acute kidney injury is overwhelmingly a result of pre-renal failure, particularly in kids. True hepatorenal syndrome is very rare in kids.
The Bits and Bobs
2 final points to consider. These patients generally have nutritional issues and this will directly impact on recovery after the operation.
Remember also that immune system thing – the risk of sepsis is much higher and will only complicate everyone’s lives.
Back to our Patient
So your 8 year old still needs an anaesthetic. She has those mechanical considerations with her respiratory function. The blood testing suggests coagulopathy, but is that really what it means? The haemodynamic state looks like the classic hyper dynamic system (though tachycardia could go with a slightly raised temperature and a bit of pain). Intravenous rehydration is under way after that vomiting.
How will you plan your anaesthetic?
Well, if you understand all that pathophysiology then the thinking anaesthetist has an expert to draw on – if you listen carefully, the patient is full of suggestions for what to do.
It turns out there is really no study to tell you which approach is best. There is no outcome study to say a particular drug combination is best.
In this surgery it’s a bit of a no-brainer, but even generally it seems prudent to have a low threshold for securing the airway with an endotracheal tube. Better ventilation characteristics might also be of benefit.
The other point for the Airway and Breathing bit? Optimise the respiratory status by choosing the appropriate positioning.
For those haemodynamics we need reliable intravenous access, boluses of fluids with assessment of responsiveness and it makes sense to reach for vasoconstrictors a little earlier than we might otherwise. Given the challenges of assessing and managing those haemodynamics you might consider invasive monitoring for cases you wouldn’t dream of usually. Like an appendicectomy even. After all, the kidneys want us to do well with the haemodynamics too.
Then there’s the bleeding stuff. Well if we’ve optimised our ability to monitor, we also want to be able to respond if there is bleeding that might be coagulopathic in nature. Pre-emptively correcting results doesn’t have a lot to back it up though. And TEG? Well there’s not the evidence yet to suggest it changes care or outcomes much outside the setting of liver transplant.
And of course, the drugs. Well there’s just a few points there:
- It makes sense to think of organ-independent options (cross reference muscle relaxants) and agents that don’t linger.
- For volatiles, outside of not using halothane if you have the option of other agents, there’s not much evidence that the choice of agent makes a huge difference. Patients with true end-stage liver disease do seem to require slightly less agent to achieve a BIS value unlikely to be associated with awareness (assuming you’re still a fan of BIS).
- When it comes to opioids, not using morphine (but using other agents) makes sense but you need to provide good analgesia. Perhaps an equally important consideration is the level of post-operative care the patient receives afterwards (somewhere that it’s easy to observe for sedation which is the big clue that a patient is at risk of negative side effects of those opioids).
- And other agents? Benzodiazepines aren’t really friendly (for example midazolam is both less effective clinically but more likely to impede recovery). Dexmedetomidine doesn’t have a lot in the literature (but note the haemodynamics again). Sugammadex seems to work but it takes longer.
- Regional anaesthesia – well who knows? I was always taught it was a worry. But then there are units who employ epidurals in liver transplant care and for hepatectomy procedures (even though we know the blood tests would suggest coagulopathy peaks at days 2 and 5). What I couldn’t find was anything in the literature relating to regional anaesthesia that reflected our new understanding of those things that promote thrombosis. Perhaps the most pertinent question with this uncertainty is “what do I gain by trying it?” How many procedures can you truly say outcomes will be greatly better with epidural?
Now, do you have a plan for our 8 year old needing her appendix out?
All the images used were on Flickr via Creative Commons and used in an unchanged manner.
Cookie Monsters was posted by Michael Verhoef.
The Lego photo was posted by David Lofink.
The Cleaner Shrimp came from Richard Ling.
Arnie was posted by RV1864.
The Parliamentary Archives is a photo from Matt Brown.
The fog was from Nikos Koutoulas.
And now the most useful references I found. First some general review articles:
Some risk stuff:
The blood stuff:
The neurocognitive stuff:
A thing on dexmedetomidine pharmacokinetics (which is open access so just go there):
Things hinting at regional anaesthesia:
Takita K, Uchida Y, Hase T, Kamiyama T, Morimoto Y. Co-existing liver disease increases the risk of postoperative thrombocytopenia in patients undergoing hepatic resection: implications for the risk of epidural hepatoma associated with the removal of an epidural catheter. J Anesth 2014;28:554-8.
Yuan FS, Ng SY, Yuen K, Lee SY, Chung AY, Poopalalingam R. Abnormal coagulation profile after hepatic resection: the effect of chronic hepatic disease and implications for epidural analgesia. J Clin Anesth. 2012;24:398-403.