BreatheTV Episode 5
SAEM 2017 | Hi-VNI Technology – A Refined Form of High Flow Nasal Cannula (HFNC)
George Dungan: Appreciate the time. I know I’m competing against dodge ball, so I have to concede a certain modicum of defeat right off the bat. Thanks for your perseverance.
I’m George Dungan. I’m the VP of Science Innovation, which in our organization is a combination of invention, medical affairs, and scientific characterization. It’s an interesting confluence. I bring that up because towards the end of this, I’m going to be talking really briefly about some of the sponsorship for research that we’re going to be initiating. I’m going to talk to you about some of those programs.
Tonight, I want to spend a little bit of time talking about high-velocity nasal insufflation, or Hi-VNI. High VNI is the trade name that it goes by. It’s a refined modality for delivering high-flow nasal cannula therapy, which I know some of you have had experience with. There are some important differences. My primary objective tonight is to help explain what high-velocity nasal insufflation is, a little bit of the mechanisms of action so that we can dive into what’s different about it, where does that exist. Then, finally, circle back around to some of the clinical application.
Just show of hands, was anybody here in Professor Doshi’s talk yesterday? A couple. Okay. I think that this will be a capstone to that, to a certain extent, to fill in some of the gaps that his 12 minutes didn’t permit him to be able to run through.
I’ve got a significant disclosure. I’m a shareholder and employee of Vapotherm. I’ll check with our CEO. Is this still an accurate slide? Okay, good. Thanks. Appreciate that.
This is the only slide I’m going to require reading tonight, so I apologize. You just have to bear with us as we go through this, but I want to start with some definitions. We’re going to circle back around to these as we move through these slides.
Hi-VNI, high-velocity nasal insufflation is a therapy that uses flow characteristics through a nasal cannula, so it’s an open system, no mask, with ideally-conditioned respiratory gas. It’s intended to accomplish two primary objectives. For inhalation, we want to approach the patient’s peak inspiratory flow. The idea there is to try and at least reduce the degree to which the influx of gas into the nares is going to entrain external gas and admix and change the FiO2.
More importantly, and frankly, if it doesn’t exceed the peak inspiratory flow, that’s probably not a problem. The more important mode that we’re trying to address is the expiratory phase. On exhalation, our goal is to provide adequate flow and specific flow characteristics. That’s really the key, to purge the nasal, aural, and the pharyngeal regions of the upper airway. In a nutshell, it’s to replace the end-expiratory gas in that extra thoracic dead space with medically-conditioned delivered gas. That’s in a nutshell what we’re trying to do.
This animation, we can use this to identify what it is we’re trying to do. This is a model that uses smoke to try and get a visual image of exactly what’s happening when we’re delivering the high-velocity nasal insufflation. As the gas enters the nares, we see a vortex form, and we see flush of this interior part of the nasal passage. That’s the gimme. That happens fairly straightforward.
The more important component is to have that gas path continue to the posterior pharyngeal space over the tongue and preferably through an open mouth. That’s actually advantageous. Unlike BiPAP, where you put chinstraps on patients to try and keep their mouths closed, open mouths are actually good here. The goal is to pretty much everything from about that point north, in that inter breath interval, we want to flush that gas. It’s a tortuous anatomy. It’s an anatomy that presents some challenges. It’s the technique of the gas flow conditioning, the velocity, that really gets us to that point.
You’re all familiar with this. Let’s just, in a nutshell, the goal that we’re trying to achieve is alveolar ventilation. This is the physiologically-required end target. The patient’s going to do whatever they have to do to try and achieve that. The work required to get to that point is a combination of things. The gating element, the determinant of that ventilatory efficiency, is the dead space. This dead-space ventilation is the piece that probably is the element that we have the most impact over.
Imagine a case where you’ve got a patient coming in in acute exacerbation of COPD, heightened work of breathing. The lever that we have access to in terms of noninvasive ventilatory support to treat these symptoms is to try and modulate this amount of dead space. That’s what this flush is intended to accomplish.
With this cartoon, red is the external, medically-treated gas. Blue represents the exhaled gas. In a normal breathing situation, the anatomical dead space serves as a reservoir for this end-expiratory gas. That then means that that’s the first cab off the rank, that’s the gas that is immediately insufflated into the lungs when I take my next breath. With high-velocity nasal insufflation, what the goal is, is to have this jet of conditioned gas actually proceed into the nares posterior pharyngeal space, and out the mouth, and do it quickly enough that we can effectively charge this extra thoracic dead space volume with preconditioned gas so that when I do go to perform that next inhalation, I’m not reducing that amount of dead space effectively.
The mechanics that we have discovered that facilitates that flush is non-trivial. There are some important mechanical differences. What it comes down to, essentially, is the caliber of the cannula that’s used. The smaller the bore, the higher the velocity. You can think of it as a finger over a hose. With higher velocity, just given a compressible fluid that’s going to increase the energy of the system, and that’s what’s important in terms of actually facilitating the flush.
This is an example of a small-bore cannula in a computational fluid dynamic model and a large-bore cannula. The color map is used to demonstrate the residual velocities of the various particle vectors. This has a much higher velocity matrix to it compared to the large-bore cannula that’s more typically used. In fact, even in this model, this is still much smaller than what’s typically used clinically.
The big effect by doing so, by having this high velocity insertion of the gas, is it dramatically reduces the clearance time. Time becomes really important, especially when you’re dealing with patients that are tachypneic. It’s that inter breath interval becomes really, really short. We’re talking tenths of a second to affect that facilitated flush.
This is the dam picture that I got from Tom. Tom keeps coming back over and over and telling me to use the dam picture because there is some dam characteristics of this that are pretty important. If you think in terms of flow, so liters per minute, gallons per hour, flow is constant in the system. The flow is exactly the same here, as it is here. There’s no difference in flow. But by choking down this broad, high-pressure back into this little, tiny orifice, I’ve now increased the velocity and the energy, more importantly, of that outflow. You can see given all the roiling of the water, and the disturbance, and the vortices, and eddies, that energy is not without impact in terms of its interaction with the rest of the fluid medium.
You can also see, and we see this in the modeling we do, you move further downstream and you can think of this as the nasal cavity that’s now dissipated. It’s to the point that it’s not creating significant deleterious impact. It’s this piece that’s the important piece for us.
When we look at typical flows that we might use, so for high-velocity nasal insufflation for an adult, we might be looking between 25 and 35 liters per minute. That’s typically more than adequate for the vast majority of the patients that we’re using. When we compare the velocities, and this is just a calculation, it’s just fluid mechanics, when we compare the velocities that we derive from those flows to the velocities that you get out of the larger-bore adult cannulas, it’s about 350% more. The flows required to achieve those velocities were getting out the flows that are just outrageous. That’s important because it’s the high velocity that facilitates the flush.
Typical flow requirements. For adults, 25 to 35 liters per minute. Again, if that doesn’t exceed the patient’s peak inspiratory flow requirement, that’s really not a problem because, again, the important aspect of this is what happens during the expiratory phase. That patient may have a peak inspiratory flow that’s higher than 35 or even 40 liters per minute, but in that inter breath interval, by facilitating the flush, I’ve now charged that extra thoracic dead space, and when the patient goes to breathe in that next breath, we’ve reduced any entrainment burden with that additional volume that’s fully-conditioned.
25 to 35 is typically where we suggest starting. In an ED, you’ve got somebody who comes in. They’re very acutely ill. You’re trying to get your arms around the patient. There’s certainly no problems starting them at 40 liters per minute and then titrating down from there. We see that anecdotally in a number of the institutions that are using our products.
Neonates I put up here. I don’t know how many neonates you see in the emergency department, but it’s an interesting case study because for neonates, high-velocity nasal insufflation flows are typically between four and eight liters per minute, which is way higher than their nominal peak inspiratory flow requirements. It’s important for them because they’ve got such a disproportionate amount of extra thoracic dead space, these little babies with these giant heads compared to the rest of their bodies. You need that higher flow to facilitate that flush in that inter breath interval.
Pediatrics. As we would say in Australia, it’s a bit of a dog’s breakfast. It’s all over the map mostly because you’re talking about patients that are a month old to 18 years old, linebackers that are my size. A rule of thumb is there’s an inflection point around six years old anatomically. This is the report for the sweet spot that has been reported to us for the high-velocity nasal insufflation. Below which, if we treat them more like the neonates, given the slight disproportion of their extra thoracic dead space, that’s probably not unreasonable. After six years, they’re probably more adult-like.
Now, people that are a lot smarter than me have helped put together a tool. The fortunate thing when it comes to pediatrics is there’s an app that exists that does the math for us. One of the things that I think is important to remember when we’re delivering this high-flow, high-velocity gas into the nose, is that it has to be properly conditioned. If we fired 25 liters a minute of dry gas into somebody’s nose, you’d hear about it from the patient, and then you’d probably hear about it from their attorney.
What we try to do is we try to heat and humidify the gas. That’s the combination. We want it to be pretty close to body temperature. Sometimes you might go a little lower, given patient comfort settings. In humidification, we really want that gas to hit the nose pretty close to 100% relative humidity.
We accomplish that using a methodology that includes a vapor transfer cartridge. If you see this cartridge cut away, there’s a number of little thin straws that are made with a porous membrane. We’ll look at the top. We’ve got all these little holes. The gas is going to be coming in from the device. We’ve adjusted the FiO2. As it passes through these tubules, water actually diffuses into the membrane. This is important because that water carries heat, and it also acts as the agent for moisture. It does so in an energetically-disintensive fashion. It allows us to have gas that exits the system in an energetically-stable, high-heat, high-humidity environment.
One of the challenges, then, is we’ve got this 100% relative-humidity, fairly-warm gas that leaves the machine. How do you maintain the integrity of that gas flow as it gets out to the patient? We solved that problem by jacketing the gas flow. It’s a triple-lumen tube, essentially. The gas flow passes through the center orifice and there’s a pathway on either side of it that passes warm water. What we’re able to do is maintain the thermal integrity of that energetically-stable gas out to the point that it leaves the end of the delivery tube and joins the cannula. The goal is to try and reduce or even eliminate condensation in the delivery tube.
Our whole, overall therapeutic goal, and this for us is a motherhood statement, is to help manage the signs and symptoms of respiratory distress, including hypercapnia, hypoxemia, and dyspnea. That’s what we do. I remember as an organization, we were bouncing around, “We’re a COPD company. Who are we?”
This is actually who we are. This is what we treat. It’s important because, especially in the emergency department context, this remit, this description, crosses a number of pathologies. We’ve found in Dr. Doshi’s study he had a fair blend of admitting diagnoses that he was treating with the high-velocity nasal insufflation.
Let’s think in terms of where this fits. This axis is a cartoon to demonstrate relative increase in patient acuity. Then with that, the level of therapeutic invasiveness that’s required to assist the ventilation, or to manage these signs and symptoms. You’ve got oxygen and non-rebreather, patient transitions, noninvasive ventilations, CPAP, BiPAP, and then eventually resulting in mechanical ventilation.
The high-flow nasal cannulas, or high-flow nasal oxygen, it’s actually really good at delivering oxygen. It does that very well. The thing that we think is a significant differentiator for the Hi-VNI, the high-velocity nasal insufflation, is actually now, and Dr. Doshi’s data, I think, really started to strongly support this in his presentation yesterday, this now extends up to help provide the same level of support in the management of those signs and symptoms that we would normally get from noninvasive ventilation. That’s, I think, the important difference between the two therapies.
When it comes to who’s the right target, this is a possible map. This is the construct that we’ve developed in terms of thinking through who’s a target for Hi-VNI therapy, for Hi-VNI. We’ll start at the middle. The axes here, we’ve got increasing work of breathing, fatigue load kind of moving across the X-axis. The Y-axis you can think of in terms of the probability that the patient’s a likely candidate for Hi-VNI therapy. The peak here is at that junction between compensated respiratory distress and uncompensated. As the patient’s blood gasses start to deviate, and some range around that, this is the likely best target for something like Hi-VNI, certainly for early experience with Hi-VNI, because your clinical context is going to help define a lot better where these lower edges are, where these upper edges are.
As we move lower in acuity, lower in the work in breathing requirement, the patient isn’t quite as fatigued, simple oxygenation might work just fine. As we move further down the line, so the patient’s becoming more fatigued, work of breathing is becoming excessive, at some point you’re going to transition and that patient’s going to require positive pressure, mechanical ventilation. It’s really understanding where the bounds are.
As I said, that’s likely to shift depending on your clinical context: nurses, respiratory therapists, the physicians ordering the care. That’s going to shift depending on the patient, too. We don’t treat the underlying condition. Just like BiPAP. We don’t treat the underlying condition. We’re just offloading the work of breathing to give you time to hydrate, and get the antibiotics on board, and do all the magic medical things you need to do to treat the patient.
This is a really important area for us. It’s important to us because if you think of the emergency department, and I’m talking to a bunch of emergency physicians, you guys know this better than me, this is the gateway to the rest of the organization in terms of the management of that acutely-distressed patient. It’s a point of confluence for those target patients that are of interest to us.
We did a pilot study. Actually, the pilot study was really important because it went a long way to help informing the construction of the study that we heard about yesterday, the five sites study that Dr. Doshi was talking about, the randomized-control trial.
I want to share an experience. This was from a case study. It was a large, regional, community medical center, emergency department. It’s Athens Regional Medical Center. This is a case study extracted from their early experience on the technology. I put it up because I think it’s informative as to a fairly-typical case. It was repeated many times during the course of this pilot study. We’ve got our normal disclaimer, so your mileage may vary. This is a case study. If I took it off there, our compliance legal team would have an aneurysm and they’d be sitting in an emergency room right now.
This is a 60-year old female patient with a history of COPD. She was a frequent flyer in this emergency department. In fact, two weeks prior to this admission, she was seen for an acute exacerbation of COPD and she was intubated. They were ready to put her on her normal conduit, which is come in through the door, go into bi-level, hope we don’t have to intubate you.
She presented with her normal constellation of signs and symptoms. They made the decision during this pilot study that they’re going to do the ABGs, and they’re going to launch Hi-VNI therapy, so the high-velocity nasal insufflation. They started at 25 liters per minute. They started at 25 liters per minute at 60% FiO2. The gasses are notable. Would you intubate somebody at those levels?
George Dungan: Not yet? Would you?
George Dungan: You’d give them some time? Well, that’s what they did. They let this patient cook along. 6:30, the admission, and 34 bicarbs, so this isn’t this patient’s first rodeo. 7:28, we see the respiratory rate start at 36. Over the intervening, or hour and 15 minutes, we see the respiratory rate drop to 22. The follow-up gasses at 7:17, 7:41, 53 for the CO2. Maintenance of PaO2 during the whole process. Maintenance of the SaO2 during the whole process.
This is the outcome that we’re trying to achieve. Managing the signs and symptoms to give you guys time to do, as I said, all the other medical magic that you need to do. They were flabbergasted at the site, just because this was a well-known, frequent patient.
When we looked at the kinds of patients that were seen during this ED pilot project, it was informative. The interesting piece from this pilot data was that the numbers are actually fairly well-imaged when we look at Dr. Doshi’s findings. It’s an interesting constellation of patients. Mostly COPD. A large population of general dyspnea, so dyspnea of unstated origin. Interestingly, a fair chunk of CHF patients that were coming through the door.
In the study that Dr. Doshi was the principal investigator on, they were looking at all-cause dyspnea with certain rational inclusion/exclusion criteria, but largely it was an all-comers study. We await those findings in their publication.
This is what we do. This is the space that we’re in. The rationale, I think, the high-velocity approach is that it treats these significant signs and symptoms. I think we’ll just stop and maybe we can just have a discussion.