Thread by @WhistlingDixie4

Up next at #LIVES2022

🧠 Optimising Brian perfusion after resuscitation.

This is another debate (read chat, more than debate)

So I’ll try to keep up…

It opens by saying we know that 2/3 of unconscious patients in ITU post cardiac arrest die

And most of them die from brain injury

(I’ll use HIE to say hypoxic ischaemic encephalopathy from now)

❓Do we need to improve perfusion to the brain and can it help anyway❓

HIE brains have inflammation, reorganisation and things which alter interaction with systemic circulation.

Robba is surprised that we don’t seem to categorise post arrest brains as other brain injuries and so treat them with the same systematic approach

Sekhon says that we tend to make the assumption that if we optimise flow and O2 content then we will have normal utilisation and diffusion - but the coupling with diffusion and utilisation isn’t in text

Mentions Box trial

nejm.org/doi/full/10.10…

So the perceived importance of improving oxygen delivery and the reality of the (lack) of outcomes means we need to rethink

And consider other parts of the picture

Skrifvars says that one group will die irrespective and a second group will do ok anyway…

But there is a third group who *could* improve with intervention and we can modify MAP or O2 or CO2 but really the issue is how do we get areas to *use* that perfusion??

(So the message now is clear - perfusion does not mean utilisation or diffusion)

But what can we do then at the bedside to look and understand if systemic targets (Like MAP 65 paO2 10) are good for the brain?

Robba says she will start far from the tech that is afforded person next to her (🇨🇦)

They use non-invasive things like pupilometry and trans cranial Doppler (TCD) to infer

Sekhon agrees they have a lot of tech - and do invasive things in 🇨🇦 but he is clear their goal is to *study* the disease

Points out he doesn’t think that’s a silver bullet - you can’t do it in all patients and he doesn’t see it becoming the gold Standard for post resus care

He mentions the jugular venous oxygenation and says that TCD is a fantastic tool

The key though is to select out the pathophysiological phenotypes and ask is there is a less invasive way to do that

Like bio-markers

(You’ve guessed it - all ICU roads lead to biomarkers)

Sekhon says NIRS is not an accurate tools here even though allure of it being non-invasive is significant

…the decoupling scuppers our evaluation (he didn’t say scuppers)

Then there’s effect of melanin on the absorption of the signal too @iceman_ex @iwashyna

So should we phenotype them and then we can pick monitoring?

We’re told maybe if significant hypoxic downtime and therefore inferred brain tissue swelling …we could lean more to invasive ways?

As opposed to patient with short time in low/no flow which wouldn’t warrant it?

Skrifvars says yes we should phenotype but need to also have clinical context (wake up test) we’re told to think of that as a biomarker too.

(I like that, I have access to it)

He agrees NIRS isn’t a silver bullet, but says in anticoagulated pts there isn’t another choice

The audience is asked if they used neuro-monitoring in post cardiac arrest care (not EEG)

The answer is very few (and I think they’re Canadian 🇨🇦)

Now to ask the panel if you had to pick TCD or NIRS which would you pick?

Robba says it depends what you want to see. She said the algorithm for TCD has stronger evidence but one is about oxygen and one is about flow so they combine to give an answer

(She didn’t pick)

Sekhon is asked by audience how can we restore O2 delivery to the neurons

He said we need to spend a year or 2 or 3 categorising problem first…thay even in TBI that target is elusive (points to work by @Menon_Cambridge)

He adds that if we find something. It won’t work across the board anyway

So when we do find something, we need ‘rational polypharmacy’ (ie a bundle of interventions)

He also points out that Brain injury is not homogenous and wonders if there is a way to undertake imaging beyond what we do now - to get better data points to inform our research

Someone from audience says he thinks only choice between TCD and NIRS is TCD because you’re measuring something directly assessed - NIRS is a dirtier signal and both arterial & venous

(I’m not sure this is a question and it is goes on longer than that but I lose track)

I’m right it wasn’t a question…

The comments though are wise, the panal say

Sekhon said that the outcomes for post arrest HIE haven’t improved like other conditions have. And we need to make the next jump….

But we don’t have a target

And we need to identify the phenotypes but again makes it clear invasive monitoring isn’t the answer

So he asks can we use biomarkers another way - as a sort of ‘brain troponin’ to trigger *something* to correct delivery

He likens it point of care testing for G-FAP in TBI

(I like sound of Brian troponin in and hope it’s more useful then a D-dimer)

pubmed.ncbi.nlm.nih.gov/32854584/

Audience Q -what do you actually do with the data you get from NIRS or TCD

Robba says they pick a MAP target (they prefer TCD) and then also estimate the ICP (with TCD or optic nerve sheath diameter)

But she stresses it’s not about exact value for ICP but if it’s high or not

And we end.
Out of time. #LIVES2022