Thread by @ebtapper

What's the deal with Ammonia in hepatic encephalopathy?
A #cirrhosis QI #tweetorial

Aims
Try to answer this question while also showing
1) Ammonia levels dont matter
2) U should not restrict protein in patients with HE

Keywords: "ammonia hypothesis", sarcopenia

First, a poll:
Does Ammonia cause hepatic encephalopathy?

How did NH3 get involved with #HE in the first place?

Step 1: In 1877 Eck develops the first successful vascular anastamosis, a portocaval shunt. ncbi.nlm.nih.gov/pubmed/9186464
Step2: Pavlov improves it
Step3: Dogs get goofy after eating meat:"Meat stupor"
Step4: What's up with meat!?

Enter McDermott and this 100% bananas paper from 1954:
ncbi.nlm.nih.gov/pmc/articles/P…

Digression:
McDermott was awesome:
War hero, fixed Muhammad Ali's hernia, revolutionized liver surgery

Here is his obit: nytimes.com/2001/07/25/us/…

So, McDermott cared for a man with painless jaundice.

In the OR:
A 5cm pancreatic cancer with superior mesenteric vein involvement.

To resect in 1 stage, he smacked the portal vein onto the inferior vena cava- Eck fistula!

He pub'd this story in 1952:
ncbi.nlm.nih.gov/pmc/articles/P…

Then the patient keeps getting readmitted for bizarre "irrational" behaviour which occasionally progresses to coma

Everyone was stumped

So they studied him to figure it out.

It took 6 months

#NoIRB

They noticed he would get confused after eating meat

Could this be "meat stupor"?
What's in meat?
Amino acids and NH3 containing stuff!

To summarize, the highlights:

First they fed him urea -->HE
(Urea --> Ammonia)
So they fed him ammonia -->HE
Thus Ammonia -->HE.
QED

Are you interested in what (specifically) ammonia does in the brain?

Lots of bad things
Let me send you to a few references:
ncbi.nlm.nih.gov/pubmed/10728803 @JHepatology
ncbi.nlm.nih.gov/pubmed/12668989 @HEP_Journal

So ammonia causes HE.

There, I said it

But is it the whole story?

Time for Jules Stahl and this 25 page single-author paper in @AnnalsofIM 1963

Turns out, Ammonia levels have nothing to do with grade of HE

We knew this in 1963!
annals.org/aim/article-ab…

Still true in 2003!
ncbi.nlm.nih.gov/pubmed/12637132
Low levels usually good but even 1 in 13 people with Grade 4 HE (coma) have stone-cold normal NH3.
And an ammonia level of 150 could mean anything from no HE to coma.

Somewhere, someone is doing sudoku with an ammonia of 210 just to prove a point

HE is a clinical diagnosis.
NH3 levels just cannot diagnose or grade HE.

But why?

Pts w/cirrhosis & portal htn circulate bacteria from the gut.

This causes inflammation

Inflammation enhances diffusion of NH3 into the brain
(review: aasldpubs.onlinelibrary.wiley.com/doi/epdf/10.10…)

Don't forget: lots of HE is PRECIPITATED by what, u guessed it, infections (inflammation!)

Cool study: Shawcross induced hi-NH3 - a la McDermott, by infusing ammonia salts
This led to 🔽 cog function
Now, Abx 🔽 inflammation
Sp they gave Abx which 🔽 inflammation
Then they🔼 NH3 ...but there was no cognitive dysfunction!
journal-of-hepatology.eu/article/S0168-…

One major reason the levels are useless:

We have no way of accounting for the impact (and amount) of inflammation.

Act 2:
Ammonia is bad and I certainly dont want to cause meat stupor!

Should I ask my patient with cirrhosis to eat a low protein diet?

Although NH3 levels are far from perfect, we definitely want to protect our brains from high NH3

If your liver is not working
and the portal HTN is shunting blood
We need a second line of defense

We have one:
It's the muscle!

Skeletal muscle accounts for >50% of NH3 metabolism in healthy people,
even more in patients with cirrhosis.

Lockwood followed radiolabelled NH3 to the muscle (JCI 1979;63);
Olde-Damink confirmed the role of muscle in many cool studies (e.g. Hepatology. 2002;36:1163)

What happens when NH3 gets to the muscle?

NH3 gets converted to glutamine at the price of a branch-chain amino acid (produced from muscle breakdown)

NH3 detox by the muscle is a catabolic process!

Restricting dietary protein is a great way to exacerbate catabolism and #sarcopenia in patients with cirrhosis

That's why patients with HE should have >1 g protein / kilogram body weight.

NH3 may rise after a protein-rich meal ... but it is safe

RCT: ncbi.nlm.nih.gov/pubmed/15246205
RCT: ncbi.nlm.nih.gov/pubmed/18627001
Guideline: ncbi.nlm.nih.gov/pubmed/16707194

So the next time you see a patient with cirrhosis are you going to check ammonia level?

And when you see them, are you going to recommend a low protein diet?

I hope you enjoyed this. Wanted to thank @tony_breu for his inspiration.

If you are interested in more stuff about ammonia, here's a plug for our paper "Refining the Ammonia Hypothesis" in @MayoProceedings sciencedirect.com/science/articl… with @Vilas_Pat and @ZGordonJiang