Many thanks to Dr. Kim for a fabulous overview of the conflicts between lung and brain protection. In this talk, I will review the ways we can monitor the brain to allow us to optimize pulmonary support without causing secondary injury to the brain. I have no disclosures. We will begin by discussing intracranial pressure. As Dr. Kim pointed out, several maneuvers performed for ARDS may elevate the ICP, but this is not necessarily a constant phenomenon. There are conflicting data on the effect of increased PEEP on ICP. Several studies have actually shown that ICP changes with PEEP can be fairly minimal and potentially clinically irrelevant, and that maybe the relationship between PEEP and ICP can change over time. This example published in 2018 of a woman with severe TBI and pulmonary contusions shows her ICP at the bottom in black and her arterial blood pressure and cerebral perfusion pressure up here. At her baseline, she was at a PEEP of 8, but was noted to be hypoxemic and adalectatic. On increasing her PEEP to 16, you can see here there's initial gradual rise in her ICP to 20, which then comes back down, and this may not have been meaningful to her. However, with prolonged maintenance of this high PEEP, you can see her ICP go up to over 25 with a further decrease in her blood pressure and CPP, suggesting that at this particular point, she can no longer compensate for the increased PEEP. So how do we know if someone will decompensate with higher PEEP? Do we just need to try it and see? Several studies have been done to evaluate which patients may be able to tolerate higher PEEP. From the pulmonary side, it appears that patients whose oxygenation responds to recruitment maneuvers can tolerate the elevated ICP without a significant change in their brain tissue oxygenation, which we'll discuss later on. This is probably because their alveoli are actually collapsed and need that extra PEEP to be recruited, rather than the patient who's already fully recruited and PEEP is just leading to overdistension of the alveoli. Additionally, patients with diminished lung compliance seem to tolerate PEEP from a neurological perspective, and that's probably because the reduced lung compliance prevents the higher PEEP from being transmitted to the thoracic cavity, and therefore has less of an impact on hemodynamics or venous return. On the cerebral side, patients with low cerebral compliance are more likely to have an ICP crisis with increased PEEP, and that also seems to make sense. Here you see the EBD waveforms on the top of a patient with normal cerebral compliance, and at the bottom, the elevated P2 here in the patient with poor cerebral compliance. If the patient is nearing the end of their intracranial compensation, small changes in PEEP may result in larger changes in ICP. So these are some ways we may be able to risk stratify patients before doing a PEEP trial. Now there may be patients where an ICP monitor is not present or contraindicated, for example, in patients with severe coagulopathy. Transcranial Doppler has been shown to be able to pick up elevations in ICP using either waveform analysis or the pulsatility index, which is grossly the differences between the peak systolic and the end diastolic flow. Remember that the brain is a low resistance circuit, so that robust blood flow can occur in diastole. Now as you see here on the left, multiple formulas have been proposed for calculating an ICP, and there does not seem to be one magic formula that correctly estimates the value of the ICP. But as the ICP goes up, you find the diastolic flow, which is here, gradually dropping, and the pulsatility index rising. You can see that in the same case that I showed you earlier, the team incinated the middle cerebral artery during the patient's baseline and with increased PEEP. And you can see a change in the waveform from the baseline, where you have normal systolic and end diastolic flow, to after the recruitment maneuver, where you have elevated systolic flow and diminished relative diastolic flow, with a more resistive appearing waveform. So this can be used at the bedside as a point of care test in the event that you do not have an invasive ICP monitor. Another non-invasive tool that can be used is the optic nerve sheath diameter. The diameter of the optic nerve sheath, measured at three millimeters behind the retina, has been shown to correlate with ICP as the intracerebral pressure is transmitted forward. And so this would be the optic nerve sheath diameter. So in the same case, here the optic nerve sheath diameter is checked at baseline, and then with a PEEP of 16. And you can see the diameter increases significantly once the higher PEEP is applied, correlating with the other non-invasive and invasive measures. Now interestingly, if higher PEEP is more likely to increase ICP, we may think that dropping PEEP might be a good idea. However, this nice study from the 1980s showed that dropping PEEP suddenly from 15 to 0 actually caused a transient ICP crisis in four patients. All four patients had abnormal cerebral compliance, again suggesting that it is when the brain can no longer compensate that these smaller changes for the lungs can cause bigger changes for the brain. And why does this happen? We see a sudden increase here in the arterial blood pressure, likely associated with a sudden increase in preload as the intrathoracic pressure drops. And in a patient with an injured and non-compliant brain, just the increased blood pressure is enough to cause a transient elevation in the ICP until the blood pressure stabilizes out and autoregulation occurs. Proning is a common maneuver used in ARDS, and again there is mixed data on the safety of it in brain injury. Some studies show that proning can improve brain tissue oxygenation even if the CPP drops, while others suggest that up to 20% of patients may have sustained increases in ICP, requiring supination. And this suggests that it may be more complicated than just ICP and cerebral perfusion pressure. We know that autoregulation is impaired after brain injury, and that hypercapnia in particular can affect vasoreactivity within the brain. Therefore, there is interest in attempting to measure cerebrovascular autoregulation in these patients. There are multiple ways to measure the cerebrovascular autoregulation, and all of these use a correlation between arterial blood pressure and various invasive or non-invasive modalities. A positive correlation suggests impaired autoregulation, as you don't want blood pressure and ICP, for example, to be going in the same direction, and a normal index of autoregulation is 0.3 or less. So if we look at the effect of PEEP on autoregulation in this study from 2013, we see that almost all patients have some amount of worsening in their autoregulation, although not necessarily all of these are clinically significant. It may also be, however, that the PRX correlation with ICP may be different than, for example, the ORX correlation with brain tissue oxygenation, and it's not clear if one is necessarily superior to the other. Invasive multimodality monitoring is used frequently in specific brain injury groups, primarily TBI and high-grade subarachnoid hemorrhage. Here you can see the collection of data that can be gleaned from a patient, from top to bottom heart rate, arterial blood pressure, calculated cerebral perfusion pressure, ICP, brain tissue oxygenation, cerebral perfusion that's measured, and then microdialysis. The brain tissue oxygen is a measure of the partial pressure of oxygen in the particular area where the probe lies, and is considered a marker of underlying metabolic state. For example, here you can see that at this point the blood pressure goes up slightly, the ICP goes up dramatically, and the CPP rises. However, the brain tissue oxygenation drops precipitously. This is an example where you may be fooled if you're only measuring the cerebral perfusion pressure. The other benefit of having a brain tissue oxygen monitor is that your acceptable oxygenation in ARDS systemically very well may not translate to an acceptable brain tissue oxygen level. In this study from 2009, you can see the initial PaO2 in TBI patients on arrival to an ICU on the x-axis, and you can see that the y-axis shows observed minus predicted survival. The range of acceptable oxygenation to improved survival is way more broad than we may think, with a minimum of 100 to 150 PaO2 showing an improved rate of survival. Certainly a PaO2 of 150 is not what we are aiming for in ARDS, and so perhaps our systemic oxygen targets are not adequate. One nice study suggested that a measurement of the brain tissue oxygen response may correlate to the PRX, and so in places where you may not have access to continuous PRX monitoring, you may be able to measure the brain tissue oxygen response, which is a comparison of the response of the brain tissue oxygen and the arterial oxygen when you increase the FiO2 to 100%. That may help you estimate whether the patient has intact autoregulation. There are more formal ways of checking for vasomotor reactivity using transcranial Doppler, including responsive mean flow velocity and the waveform to CO2 administration, breath holding, and acetazolamide, or in the opposite direction, hyperventilation, but obviously these can be quite cumbersome in a patient who has poor ventilatory reserve. NIRS is a non-invasive way to assess brain oxygenation approximately two centimeters from the scalp, and therefore can be used for local oxygen monitoring. In this study, you can see what happened in patients with ARDS in response to various maneuvers. You can see that both recruitment with PEEP and initiation of ECMO actually caused immediate decreases in brain oxygenation as measured by NIRS, despite systemic improvement in oxygen. This is probably because, at least in the case of ECMO, the rapid decrease in CO2 actually caused vasoconstriction, whereas initiation of inhaled nitric or proning actually resulted in improvements in brain oxygenation. Of course, one limitation of this particular modality is that it only gives you local oxygenation in the bilateral frontal lobes, and this may not be useful in patients with frontal contusions, pneumocephalus, or frontal strokes. A more global way to monitor brain metabolism in a non-invasive way is with jugular bulb oximetry, where the oxygen content difference between the arterial blood and the venous blood within the jugular bulb, right here, is measured. In this nice study, you can see that patients were recruited from 0 to 5 to 10 of PEEP, and patients seemed to divide into two groups. Ones who were considered recruiters, who were responsive, both systemically, to increases in PEEP without significant decreases in their oxygenation within the jugular bulb, and then non-recruiters, who did not have significant improvement in their oxygenation and had worsening jugular bulb numbers. This suggests, again, that those patients who need recruitment due to collapse of alveoli may be able to tolerate higher PEEP, whereas those patients where increasing PEEP is just over distending the alveoli will have that positive pressure transmitted to the thoracic cavity and therefore result in harmful effects to the brain. Thank you for your attention, and with that I'll turn it over to Dr. Desai to discuss some cases.

lung and brain protection

acute respiratory distress syndrome

intracranial pressure

brain tissue oxygenation

vasomotor reactivity