Conference sponsored by the Extracorporeal Life Support Organization (ECLS) in Ann Arbor, MI on 8-9 March, 1996.
I felt immense frustration at Alliance Pharmaceuticals (the developers of perfluburon (LiquiVent, TM), the working fluid now under Phase II/III FDA trials) for their lack of communication with interested investigators (including me) and for their lack of willingness to make LiquiVent available to many (even most!) of the most respected, conservative and highly qualified investigators who requested it for *animal* research (definitely NOT me!).
REGARDING SOME OF THE CAVEATS:
Since I am breaking this review up into parts, I want to state at the outset that my caveats re: Alliance and the near-term utility of liquid ventilation have been greatly modified. I will discuss the latter in due course, but feel compelled to say that Dr. Mark Wedel of Alliance was responsible for a major attitude adjustment towards Alliance on my part. I think Mark is a rare bird in any flock of people pursuing new technologies, and in particular in companies pursuing new medical biomedical technologies in the land of hype: my impression is he is a straight-forward and no-bull kind of guy. I got clear, blunt, and I am quite sure ;)) *honest* answers to some very sensitive questions I asked him privately. I saw him handle others, and his presentation on LiquiVent, similarly.
IMHO Alliance is fortunate to have Dr. Wedel associated with, and speaking for them, and they could profit greatly by getting others with a similar degree of competence, directness and courtesy in their corporate offices in San Diego, and particularly in their interface with investigators interested in both animal and human trials of their product.
Finally, I met some other CCM-L people there, and had several cordial and informative conversations with Dr. Jeff Rosenberg of Johns Hopkins. I also met with a couple of CCM-L critical care physicians who drove an hour from a nearby city to have dinner with me and discuss resuscitation research. I thank them for a stimulating evening and for their willingness to spend their valuable time with a "fringe-type" (FL's words for me) like me. To protect their reputations, I'll not mention their names!
ELSO CONFERENCE REVIEW
The conference opened with a presentation on ARDS Pathophysiology and Physiologic Management by Robert H. Bartlett, Professor of Surgery and Director of Surgical Critical Care, University of Michigan Medical School. If anyone expected this to be a snore or a basic review, they were in for a surprise. In fact, Dr. Bartlett's positions on several issues were being discussed heatedly even after the conference was over on Sunday I shared a cab back to the hotel with two intensivists who had strong opinions of their own and who were asking *how* Bartlett could do what he claimed.
Perhaps the most controversial element of Dr. Bartlett's management strategy was his emphasis on the importance of aggressively treating *peripheral* edema (fluid overload) in ARDS. Dr. Bartlett was quite firm that this strategy should be followed for several reasons:
Of course, HOW to achieve this end *and* maintain hemodynamic stability was the topic of debate. What about optimum preload? How do you keep PAs and thus CO in the acceptable range if you can't "tank-up" your patient with fluid? Bartlett's answer was, in part, to transfuse to a normal hematocrit (none of this "leave 'em at 30 -- it improves hemodynamics") *and keep the patient there!* He argues that normal to up-end-normal crits are the most powerful way to increase VO2. A side benefit is that the RBCs represent a" leakage-proof" source of vascular colloid -- colloid which will help keep water in the vascular space instead of allowing it to leak into the interstitium.
He also pointed out that renal failure as a consequence of such intervention is *not as critical as lung failure* and, while to be avoided if possible, was certainly a better choice than ARDS or exacerbation of ARDS. As Bartlett pointed out, this attitude of accepting renal failure with less panic is now possible because of the relative ease with which this complication is now treatable (CAVHF...) and the virtually universally transient nature of the renal failure, i.e., reversible ATN versus permanent injury to the kidney. He said that the longest they had a patient in renal shutdown in ARDS was (if I remember correctly) in the range of 28 days -- not typical for ATN in my experience. His second major thrust in ARDS patient management is one I could not agree with more: NUTRITION, NUTRITION, NUTRITION! Early nutrition. Immediate nutrition. Nutrition in the face of rising BUNs. Nutrition enteral first and whenever possible, but via TPN if necessary without hesitation, and early, rather than late.
His points here are often lost on many intensivists and critical physicians:
*The catabolic patient has enormous nitrogen, vitamin and trace mineral demands.
*Tissue repair is urgently needed and lung repair, other organ system repair, and maintenance of host defense (immune function) cannot be carried out well, if at all, in the starving patient. Critically ill patients, with ARDS or not, need immediate nutritional support. As one conference attender sitting near me remarked under his breath (Jesus! This guy talks like an Internist, not a surgeon!).
* Azotemia can be managed with hemofiltration, dialysis, etc. Several other points Bartlett made during his presentation which are worth pointing out:
* The consolidated ARDS lung is a lung in which a substantial portion of the lung is unavailable not only to gas exchange, but to *gas filling*. This point was first made (as Bartlett noted) by Gattinoni some years ago. Gattinoni was also the first to realize the corollary of this: continued use of "normal" tidal volumes in the ARDS lung does not result in inflation of the consolidated lung areas, but rather in over-inflation and *added injury* from volutrauma and barotrauma to the remaining gas accessible, relatively uninjured lung. Thus, Bartlett strongly argued for pressure cycled ventilation and for the use of moderate peak pressures over the volume-cycled, high pressure approach. His explanations, anecdotes and accompanying graphics were lucid, amusing and comprehensive.
Anyone who has autopsied humans or animals with ARDS will know the truth of what he said: as we in our lab (and as another CCM-L contributor pointed out here a day or two ago): there is nothing quite! as educational in this regard as opening the thorax and seeing two "livers" without a bubble of gas in them, with tiny slivers of normal lung tissue sitting atop them, to appreciate just how true this is. Thus, as Bartlett pointed out per Gattinoni, what we have is "little baby lungs to ventilate" even in a large adult. Ramming away with the same tidal volume just recruits more alveoli into failure and consolidation.
A last point Bartlett was at pains to make was the tremendous importance, in his opinion, of prone positioning in improving gas exchange and outcome in ARDS. He had very dramatic CTs of migration of the consolidation from one dependant area to the other over a period of several hours. Their (UMMC) protocol calls for prone positioning of ARDS patients with a protocol of 6 hours prone and 6 hours supine, using one of the automated air-cushion beds to carry out routine side to side turning, percussion, and avoid pressure sores. Bartlett had a number of slides dealing with this topic including a nice one which showed TLC, FRC and RV versus various body positions from supine through sitting, standing and prone. Not surprisingly, short of standing the patient bolt upright, prone positioning with pressure taken off the abdomen (they use rolled up bath blankets, towels, etc to keep the abdomen off the bed surface) results in the most physiologically optimum of all three parameters.!
Supine positioning allows the abdominal viscera to rest on and compress the distal and most dependant (and best perfused!) lung exacerbating v/q mismatch. There were a number of questions about the logistics of this: especially with 150+ kg patients. Bartlett stated emphatically that they had managed prone positioning with a minimum of problems even in the most obese patients and that he considered it an essential ingredient in successful management of ARDS. Certainly anyone who has looked at Gattinoni's CTs from the early 90's cannot fail to be impressed with this approach at least on the basis of its undisputable effects on pulmonary parenchymal/alveolar fluid distribution (see Gattinoni, et al., Anesthesiology 1991;74::15-29).
Bartlett's talk was very information-dense and I found it immensely useful. I can only hit the high points here, but can also refer you to a concise, affordable, opinionated (and likely to be very controversial) summation of his critical care algorithms in the form Critical Care Physiology (ISBN# 0-316-08269-4) and its companion handbook of charts and tables for bedside use, The Michigan Critical Care Handbook (ISBN# 0-316-08268-6) by Robert H. Bartlett, Little Brown and Co., Boston, 1996. Apparently I was not alone in this estimation. ALL of the local Ann Arbor bookstores which had this book in stock were sold out by noon following Bartlett's presentation with extra copies to be Fed-Exed in on Saturday (not possible). I got my own copy of both books only by the expedient of immediately calling one of bookstores and reserving copies with a credit card. Even that would not have worked but for the generosity of a fellow CCM-Ler who heard the salesman take my call and decided to have his purchase shipped so I could carry my copy back. Thank you! I was able to read it on the flight back to California.
Whether you love some of his positions, or hate them, I think you will find both these volumes indispensable sources of useful, user-friendly information (the graphics and care algorithms are excellent and beat trying to use tabular or text sources for this information, hands down). The Handbook is laminated and lab-coat pocket sized. Finally, a decidedly non-objective impression. Bartlett is a commanding presence and formidable intellect of the old school mold. He is obviously greatly respected as a surgeon by his colleagues. I would not like to cross swords with him, right or wrong. Which leads to a disclaimer: if I have misrepresented Dr. Bartlett's position of statements the error is mine, and I feel sure I will be corrected ;).
Some further thoughts on Dr. Bartlett's presentation on management of ARDS: As I noted in Part II of this review, Bartlett advocated aggressive fluid removal to dry weight and support of HCT. Removal of even 100 or 200 cc of pulmonary water will result in the recruitment of a clinically significant number of functional alveloi. As Simmons, et al., point out, a fluid loss of >3kg resulted in 33% mortality, while a fluid gain of <3 kg resulted in 100% mortality in ARDS patients (Simmons, et al., AARD 1987;135:924-9).
Bartlett advocates use of VO2 Sat as the principal yardstick for evaluating the efficacy of perfusion and gas exchange. As he correctly points out, the use of ABGs to determine perfusion status is frequently of little utility in sepsis, ARDS and CO poisoning. Indeed, high O2 sats in arterial blood can be very misleading in the presence of low Hgb and the best guide to patient status is the mixed venous saturation. (And, to belabor the point, "low" Hgb levels can only be determined in the context of the patient's correctly evaluated oxygen consumption needs.)
Bartlett also suggests the very rational (but how widely used?) approach of optimizing PEEP by raising it until VO2 begins to decline. Forget about the paO2, go by the venous sat. Obviously the most practical way to do this is with a continuous VO2 monitoring Swan, or pulse ox (if you are poor like me, or work for an HMO :0. Heh, anybody got a fiberoptic VO2 System they wanna sell cheap?)
By using this approach, coupled with daily metabolic monitoring, including 24 hour urines for nitrogen balance assessment, and aggressive use of the aforementioned therapeutic modalities (by titrating therapies to the patient's needs) Bartlett emphasized that the need for multiple, daily, ABGs, electrolyte panels, and other costly ICU tests could be greatly reduced. In fact, he stated that their current average number of ABGs per ARDS patient per day was now down to 1.0 or 1.2 (this elicited some disbelief from some conference participants during casual conversation on the tour of the hospital later in the day). The cost savings from reducing the need for such testing reportedly more than offset the cost of the once-daily comprehensive metabolic evaluation performed at ca. 0500, and which, in conjunction with continuous VO2 monitoring, is subsequently used to determine need for nutrition, hydration, PEEP, FiO2...
One last, elementary, but not unimportant point made by Bartlett, which I at least had not previously considered, was the role of N2 in the breathing air in maintaining alveolar patency. As Bartlett points out, at very high FiO2s poorly ventilated alevoli will be collapsed by the nearly complete removal of oxygen by the desaturated and nitrogen-free RV blood. FiO2s below 6.0 result in nitrogen depletion of alveoli (and tissues and blood) to the point that many alveoli will become atelectatic during exhalation (absorbition atelectasis). This means less gas exchange and probably extra injury during reinflation. Bartlett seemed to be saying that PEEP at high FiO2s is not effective at maintaining alveloar patency, or not as effective as PEEP plus FiO2s low enough to leave the N2 concentration at a reasonable level. Made sense to me.
Pne reason why mortalities remain so stubbornly high in ARDS is that people are focused on looking for the "magic bullet." This prevents them from seeing small improvements and *building* on those improvements by combining promising modalities and keeping careful records of outcomes. Raising HCT, judicious use of hypothermia and paralysis, careful attention to maintaining a positive nitrogen balance, none of these alone is likely to constitute a huge factor in improving outcome. But, alas, as in cerebral resuscitation (in our hands) the devil is in the detail and combining modalities can give not only additive, but synergestic effects. If UMMC's stats on survival in ARDS vs. the national average are any indication, this approach to managing the details, and supporting the patient's own self-repair capacity optimally, are paying big dividends.)
Next up was Ronald B. Hischl, M.D. of the University of Michigan Department of Pediatric Surgery to present the "Basics of Liquid Ventilation." He began by discussing the structure of perfluburon, the gas transfer compound they are evaluating for LV at UMMC. Perfluburon (PFB) is an 8-carbon chain molecule with a bromine atom on one end and 17 flurorine atoms "substituting" for carbons (C8,F17,Br) manufactured by Alliance Pharmaceuticals of San Diego, CA. We were provided with ranges for pertinent properties of liquid ventilating media in general (i.e., vapor pressure, diffusion coeffecient, surface tension...), but not specfic numbers for PFB, something I was probably alone in finding annoying. Suffice it to say that all the subsequent indirect evidence is that PFB meets the criteria for the "ideal" working medium:
*Low surface tension
*High spreading coeffecient
*Low viscosity (*very* critical)
*Mid-range vapor pressure (you don't want it to boil off right away, on the other hand you don't want it in the recesses of the lung 10 years later, either) *High CO2 dissolving capability (a *major* limiter on all liquid breathing media)
*Good oxygen carrying capacity * Chemical and biological inertness * Water and lipid insolubility * Per inertness above, no toxicity Well PFB/LiquiVent fills the bill on all of these. And more. And its the"and more" part I'll be spending a substantial amount of time talking about later. So, now to the BIG question, why in the hell would any sane person even think of filling up an ARDS patient's lungs (or ANY patients' for that matter?) with a *liquid* industrial chemical that would seem more properly suited for coating frying pans or making fat-free potato chips? Well, the answer is complex. So, pay attention.
We all know that water is a lousy gas exchange medium. Nature knows that too and that's why it went to all the trouble of making hemoglobin (and its copper cousins) and then packaging the nasty, reactive stuff (it's a wonderful NO sink!) in RBCs. Iron carries air, or more precisely, O2 and CO2. In ARDS we have water, an undesireable liquid occupying a large and often lethal fraction of the *best* perfused lung, resulting major V/Q mistmatch. But, the problem goes further. The water has a high surface tension, is contaminated with foaming agents (proteins) acts as a wonderful solvent for alveolar surfactant, is a nice sink for the hypohalous acids generated by granulocytes, and probably serves as wonderful propagating medium for free radicals of all types.
The high surface tension and the presence of protein result in foam and air lock, and those of us who have ever tried to manage fulminating pulmonary edema will appreciate the barrier to gas exchange *foam* represents. I have had some success in treating this in my human cryopatients by using nebulized antifoam-A (a silicone compound made by Dow Chemical) dissolved in 30% ethanol and water *at the start* of extended CPR. But such maneuvers are only a stop gap, and do nothing about the water-logged, consolidated, V/Q mismatched lung seen in ARDS. Further, in patients with immature alveoli, or neonates suffering from congential diaphragmatic hernia, the problem is collapsed alveoli which are refractory to PPV with or without PEEP.
While prone positioning, suction, elevating FiO2, increasing COP and HCT can all be of some help, the problem remains that we cannot get the water out of the alveoli or out of the lung parenchyma. We can't suction alveoli, we can't prevent their rapid refilling, we can't prevent lung parenchymal edema, we can't deal with the foam, *and* we can't deal with the micro-obstructive proteinacious and mucopolysaccharide debris that would rapidly accumulate even if we could flood the compromised alveoli with gas.
However, IF we had a liquid that was denser than water (preferably twice as dense, as LiquiVent is) and if that liquid did not foam, was inert, was a good gas exchange agent, and had a low enough viscosity, we might have a powerful tool to treat ARDS with. In fact, this is exactly what LiquiVent seems to be. When it is instilled into the lung it rapidly displaces water from the dependant alveoli. It does not foam and it has a superb spreading coeffecient and low viscosity. Due to its fluorination it is radiopque, which is both an advantage (mostly) and a disadvantage. Since it is twice as dense as water it acts as a localized PEEP. It carriers O2 at a rate 20 times that of saline and handles CO2 removal reasonably well (working solubilty is about (I'm guessing) 10-15 m CO2l/dl).
If this were all it did, it would be plenty. But it has other advantages as well. Because it displaces alveolar water and alveolar air, it acts as an alveloar lavaging agent. This allows proteinacious and mucus debris to be washed from the alveoli on each ventilation. Further, because this debris is not soluable in LiquiVent it does not cause any alteration in the viscosity of the agent or its fluidity (no "gelling" or "foaming"). A point *not* made by Dr. Hirschl or others is the inability of LiquiVent to support or communicate free radicals and toxic materials. In the normal picture of ARDS you have wet parenchyma and fluid-filled alveoli. These fluids are NOT inert, they are full of reactive species, proteolytic enzymes, free iron, and just about every other nasty thing you can imagine. And, thesenasty things are optimized to diffuse rapidly through water and react in water. Thus, the consolidated alveoli in ARDS (IMHO) serve as cesspools of damaging chemistry.
LiquiVent reduces or eliminates those cesspools and it acts as a micro lavaging agent removing mucus and debris which are also loaded with proteases and myleoperoxidases. Remember, purulent mucus is green because of the CHLORINE because of the presence of myleoperoxidase -- a product of leukocyte activity which is greatly up-regulated in ARDS and MSOF.
So to summarise, why liquid in the lungs in ARDS? Because it:
* recruits alveoli to gas exchange
* displaces toxin laden alveolar water
* elinates the air-water intrerface with accompanying elimation of "air locking" and foaming.
* provides good gas exchange
* provides more or less continuous alveolar lavage
* provides gas exchage selectively to the most dependant, most consolidated, and best perfused portions of the lung (and thus the most V/Q mismatched).
So we now have the "why," what about the "how?" Now, this where I got surprised when I first realized that others were doing LV research (about a year ago). The concept of using a liquid to completely replace lung gas space (total liquid ventilation, TLV) is pretty straightforward given the above considerations and the availability of a compound like perfluburon. But that is not the clinical approach being most widely evaluated now. Rather, what is being done is to replace only the functional residual capacity (FRC) of the patient with LiquiVent and then carry on with conventional mechanical ventilation as usual.
Say what!!!!!!!! Yup. In the Phase I trials they simply unhooked the vent line and poured the LiquiVent right down the patient's ET tube from a beaker until the calculated FRC was reached. In order to avoid vapor locking, patient discomfort, and transient hypoxia from large airway obstruction (until the LiquiVent gets distributed to the alveoli) they have since gone to adding the agent through a port on a connector placed between the vent circuit and the ET tube in roughly 60 cc aliquots. The fluid is added UNTIL A MENISCUS APPEARS IN THE ET TUBE!!!!!!!
Yup! Mechanical positive pressure ventilation is then carried out as per usual. When we went on tour and saw a patient on partial liquid ventilation (PLV) they were using a standard pediatric ventilator. For most of their adult patients they use the old workhorses, the PB 7200s.
How do they DO this, you ask? Well, it's simple really, as are all truly beautiful ideas; the LiquiVent surges back and forth between the alveoli and the larger airways. The larger airways (down some unspecified distance from the carina) are filled with vent gas. There was a lovely video of a bronchoscopy of a PLV patient: what you see is foam free LiquiVent surging into the large airways on exhalation and percolating with vent gas during exhalation and inhalation where, presumably, gas exchange occurs. Thus, the gaseous tidal voulme is the same and the large airways serve as a turbulent chamber where gas exchange occurs. Does it work?
Now, some corrections, Travel-Royce Johnson pointed out to me that: It's *S*vO2, (mixed venous saturation), I think you mean, not VO2 (the oxygen >consumption). He is correct. I apologize for the error.
Secondly, in rereading my post per Travel-Royce above, I realized that I was less than clear about Gattinoni's work and its implications for ARDS management, the biomechanics of prone positioning, and a major reason why LiquiVent works as it does. The reason that the dependant areas of the lung become consolidated is *not* due to the shift of fluid down through the tissue under the force of gravity. The rapidity of the movement or reappearance of the consolidated area at the new site of dependance in the lung on CT after prone positioning rules this out. Furthermore, I forgot to tell of Gattinoni's most elegant method of elucidating what was happening: he calculated the Hounsefield number of ARDS lung tissue versus normal lung tissue using CT. He showed that the water distribution in ARDS lung was essentially the *same* in both gas-filled and consolidated ARDS lung, and much greater in both of these areas of ARDS lung than in normal lung.
So, what is going here is that the *weight* of the wet, non dependent lung is compressing the dependant lung. A good analogy would be to envision two boxes full of balloons. In the first box the wall thickness of the balloons is a couple of mil of rubber. In the second box the wall thickness is a 10 mm of lead-doped soft rubber. All the balloons in both boxes have a tiny pinhole leak in them. What will happen to the balloons in the second box?
Clearly, the ones the bottom of the second box will be taking a massive load from the ones above (compared to those in the bottom of the first box). They will empty first and much more rapidly than those in the first box, *or those at the top of the second box.* This is why dependant lung is consolidated. Once you understand that, you are in good position to understand the beauty of LiquiVent. It is far more DENSE than water and gas. And it carries O2 and CO2. Thus, it can "prop up" the dependant alveoli or "squeeze out" water from the dependant parenchyma.
This is what has been tried with the use of surfactant in ARDS. In fact, if you look at Guyton's Textbook of Medical Physiology carefully, you will discover that the unique arrangement of surfactant molecules in the normal alveolus provides a substantial amount of alveolar rigidity and resistance to collapse. So does the normal lung alveolar capillary network, which is also degraded in ARDS (you can count on major increased activity of collagenases and elastases to degrade basement membranes, and phospholipases to degrade endothelial integrity). Unfortunately, surfactant is water soluble and, in any event, can't reach the areas where it is needed. Furthermore, surfactant probably can't situate itself properly on the compromised alveolar surface, even if it is delivered there. It was not designed to work under such conditions. PEEP will be effective only insofar as it can LIFT the weight of the column of wet lung over the dependant area. Unfortunately, it takes much more than 40 cm of water to lift and inflate the dependant lung areas. Those balloons at the top of the box have lead-filled walls! I apologize for my lack of clarity on this issue; it is critical to understanding how LiquiVent works and it was a disservice to the elegance of Gattinoni's insights and genius.
Now, back to the review of Hirschl's presentation. What kind of results are they getting with LiquiVent? Hirschl started with a slide showing FRC Gain in gas vs. (PLV) Partial Liquid Ventilation and Total Liquid Ventilation (TLV) treated patients. I made a quick sketch of this slide and my sketch shows a scale of 0-4 with PLV and TLV being increased over gas by 3-4 times. 30 minutes after PLV is started, FRC is down, but there is rapid improvement in PFC and gas exchange. With continued PLV support there is *75% improvement in FRC.*
However, the numbers are not the most stunning thing about PLV with LiquiVent. The radiography is! There is no substitute for seeing for yourself. Trot down to the library and look up "Partial Liquid Ventilation with Perfluburon during Extracorporeal Life Support in Adults: Radiographic Appearance" by Kazerooni, et al. (Radiology 1996;187:137-42). Look at the distribution of this stuff on X-ray after 1 hour! The mechanically accessible areas of the lung (including and *especially*) the dependant consolidated areas are "whited out" with LiquiVent. Stunning! Absolutely stunning. I can think of a couple of very pompous radiologists I've known who I'd love to sent such CXR to in-hospital with a request for interpretation and a note that the patient was doing well and had excellent ABG's and SVO2s ;).
Which brings us to some more specifics. The dose being used clinically appears to be in the range of 5-80 mL/Kg. My notes indicate that Hirschl remarked that the maximum dose of LiquiVent used clinically was 60-80 ml/kg and the optimal dose established in animal models of ARDS appeared to be closer to 90 ml/kg. Mechanisms of action, were stated by Hirschl to be:
Some useful references which readers might want to look at are (in no particular order):
Leach CL, Crit Care Med 1993; 21:1270-78 (A prospective randomized controlled study of the effectiveness of PFC in gas exchange).
Schaffer TH, Pediat. Res. 1978;12:695-698. (Early conceptual paper)
Furhman BP, Crit Care Med 1991;19:712-22 (Details kinetics of gas exchange with PFCs)
Hirschl RB, Crit Care Med 1995;23:157-163. (Design of a liquid ventilator).
Hirschl, RB, Chest 1995;108:500-08 (*Recent* overview of outcomes and effectiveness of ventilation with PFB in ARDS.)
Some other points Hirschl made:
15% of VO2 can be supported via peritoneal circulation of PFCs! * A 3% "flurocrit" emulsion (LiquiVent? Other PFC?) can support 50% of DO2. Since the micelles are 1 micron diameter on average, its distribution and rheology are better. There is a VAST literature on the Green Cross Company's Flusol and Flusol DA products (a binary IV PFC product) which documents its use in every setting imaginable. If I had ever been able to get my hands on the stuff in sufficient quantity, there would even be a paper on its uses in oxygenating the dead, which would have completed the range of settings/ conditions in which it has been evaluated! Flusol probably has similar intravascular rheology and "perfusability" advantages in common with emulsified PFB.
Hirschl noted that Alliance was also in process of developing a PFB blood substitute and was aggressively investigating use of this agent in organ preservation. (I will note that there are several papers on IV PFB emulsions, but these can best be uncovered by a literature search tailored to your own needs and interests.)
Some Observations and Opinions on Hirschl's presentation: This was good presentation overall. But there were several problems not just with the presentation but with the whole way liquid ventilation was presented. One of the things that left me and MOST of the ELSO conference attenders bewildered was the lack of context for all this. We humans are story-creatures and we can't make things "fit" easily into our world-view if they don't have context. I shook my head after the conference with disbelief that no one bothered or maybe thought to offer a brief history or overview of the development of LVT. At first I thought, "God! I really blew it on this one. It is obvious everybody here knows all about this and that an historical overview would be about as appropriate as talking about Gibbon and the history of the heart-lung machine and evolution of the technology of bypass during the ECMO part of the conference. These people all KNOW the score. How far behind the power curve am I, anyway?"
Well, I AM dreadfully behind the power curve, but so were most others. And, at least some of them had the same questions I. You are talking about a completely new, avant garde technology. The conference organizers might have profited from putting it in context and explaining how it came about, and who the *people* were/are behind it and what they did. As it was, Alliance was just this faceless, personality-less THING. Further, the lack of clear delineation of the relationship between Alliance and the Univ. Michigan Med. Ctr. (UMCC) investigators was problematic. Some conference attenders were very suspicious of Alliance's credibility and of the credibility of the UMMC's work. The absence of investigators exploring clinical and laboratory investigations of LiquiVent from other institutions (understandable in the context of the ELSO-sponsored conference) did not help. This could have been remedied in part by referring to other studies and indicating which other centers and investigators were specifically involved in evaluation.
In Part IV of this review I commented on dosing with LiquiVent (ml/kg) per my notes from Hirschl's first presentation. I was unable to ask follow-up questions after his presentations. I had to leave the conference (which over-ran by about 20 minutes) to make a 1700 appointment elsewhere, so I never got ask about the details of dosing and, more to the point, how they had arrived at the optimum dose. This was later clearer to me in reading papers after the conference. I'm trying here to stick to what I heard said (or *thought* I heard said). I welcome comments, expansions, or corrections from others. In the case of the dosing issue, I received an authoritative post (privately) from another of the conference participants. She corrected me (clarified my statements?) on this point, and with a little editing to change *style* not content I have posted her comments:
In all the clinical trials I am aware of, dosing is stopped when (a) a meniscus of fluid is observed in the ET tube off PEEP, or, alternatively, (b) when the dose has reached the predicted NORMAL FRC for the animal or human involved. In humans, that's 30cc/kg, rabbits 18 cc, etc. Clearly there can be an objection regarding using a reference point of NORMAL FRC in a diseased lung, however, this endpoint is driven not by efficacy but rather by safety (FDA mandated in the trials). What Hirschl was trying to demonstrate in his dose response curves was a phenomenon IDENTICAL to the original Peter Suter, Barry, Fairley "best PEEP" paper in the New Engl J Med in the '70's, namely, that ultimately too much of a good thing impairs venous return and as a result, diminished O2 delivery. In the case of PEEP the point where "good turned to bad" was called "Best PEEP." In the case of perflubron, that same point just happens to be the predicted normal FRC for that species.
Hence, the numbers. I, too, found Hirschl's articulation of the idea a bit confusing when that data was presented. I've been following liquid ventilation research for sometime now, certainly before most of you at the conference, and this, combined with being significantly older than most of you, has given me the opportunity to see the correlation with the very first "best PEEP" paper. Few in that audience were old enough to remember that.
To summarize, if you talk with the investigators or the people from Alliance, you will find that 30cc/kg is the MAX for human dosing. I understand that there is pre-clinical data which suggests that superseding this dose may well result in adverse effects on venous return.
I will review the next two presentations together. I would caution that I may be very confused on who said what. Both were information-dense and I don't note a clear delineation between the two presentations.
These presentations were "Laboratory Studies and Applications" by Paul Gauger, M.D., General Surgery, University of Michigan, and "Lung Protective Effects" by John Ender, M.D. (I believe Ender was substituting for John Younger, M.D.) and I did not note Dr. Ender's affiliation, and he was not listed in the directory of presenters provided in Conference Attenders Packet.
Gauger reviewed the laboratory studies covering a wide range of models and including both PLV and TLV. I was quite surprised at not only the breadth of the laboratory investigations, but also their depth. As per the quoted correction above, these people know a great deal about the effects of LV on hemodynamics, gas exchange, the immune-inflammatory response, alveolar development in the LV-ed premature animal, among others.
I could spend pages just restating his presentation and commenting on this work. Instead, I'll try for a brief summary and suggest the obvious: those of you who are interested should comb the literature. Tips for doing this are to use key words in your searches such as perflurocarbons, ventilation, lung, perflubron, Remar, perflurodecalin...Another excellent approach is to find a few of the papers I cited before and then do searches using the authors' names. This is, in fact, the most effective way I've found so far to find information in this area. I would encourage UMMC, Alliance, or some other institution engaged in this work (or perhaps an outsider would be better!) to establish a Website and a comprehensive bibliography. One of the first things I did when I learned of Alliance's existence was to call and ask for a bibliography, ask if a database or Website was available, and ask for an Annual Corporate Report and any promotional materials or materials for prospective investors. None of this was available. From this I must assume that Alliance has all the capital they can possibly use, intellectual and otherwise; an enviable position indeed! (I am NOT being facetious here: this can really be the case; sometimes more money, brains, and ideas just cause problems once you are saturated to the limit of! your ability to deal with what you've got).
I'll briefly synopsize my *impressions* of Gauger's presentation: LiquiVent is extremely effective in animal ARDS, reducing mortality from 40-50% to 10-15% -- or lower. It is as effective, or more so, as ECMO in neonatal RDS due to prematurity, meconium aspiration, and simulated congenital diaphragmatic hernia. Long periods of PLV have been well tolerated in all animal models explored. And, pay attention here, models of induced community acquired pneumonia showed a big reduction ********** in mortality?***** in PLV treated vs. Positive Pressure Ventilation (PPV) treated animals. Oleic acid and saline lavage or combinations thereof were used to induce ARDS in pigs and sheep. The reduction in mortality was very impressive (about 80% survival vs. 20-40% in controls).
More interesting were the "preliminary indications" (supported by subsequent clinical presentations) that LiquiVent may represent a revolution in neonatal RDS care by virtue of its ability to limit injury normally associated with PPV and to *foster continued, normal morphological development and generation of alveoli in the immature lung.* In other words, PLV seems to stop the fibrosis, honeycombing, alveolar loss and long term loss of vital capacity often seen in the most severely RDS afflicted neonates (adults too!). Anyone with medical acumen who watches television will have seen these "marginal" rescues in their later years with a portable liquid oxygen unit, nighttime vent, and other astronomically costly and damnably inconvenient paraphernalia, all required to skate these youngsters along on mostly fried lung. LiquiVent seems to inhibit or greatly reduce this kind of injury and to allow really undeveloped lung to proceed with normal alveolar development without inflammation, pneumothorax, and all the other problems that attend PPV in really bad neonatal RDS.
Particularly impressive were the CXRs of the profoundly compromised and underdeveloped lung of congenital diaphragmatic hernia. The films and histology were indicative of not only "inflation" of the collapsed lung, but of progression of normal or more nearly normal alveolar development and generation. If this finding is real, it could be of tremendous importance. I also suspect that if this work holds up in human clinical application it will allow for rescue of many babies now simply considered nonviable, indeed, considered fetuses today.
Gauger also pointed out the increased simplicity of PLV over ECMO. This is not an idle point. ECMO requires highly skilled personnel, incredibly costly capital equipment, a major institutional commitment to providing floor-space, surgical expertise, costly in-house and on-call people, and on and on. There are a million things that can go wrong and there are currently no automated process control systems clinically available (:)))) to make the treatment safer and more user friendly (the primary reasons for this are lack of economy of scale (too few cases) and the FDA which now regulates devices and makes introduction of new equipment for small-volume or start-up technologies prohibitively expensive. As to ECMO's heartaches and challenges, I *know,* I do it! (Sadly, my "dead" cryopatients get better care in this respect than they do when alive. I can use any drug or device I want: they are legally (small "d") dead if not biomedically DEAD. More on this when I cover the! tour of the hospital ECMO facilities.)
Gauger's bottom line here, while not explicitly stated, was clear. PLV means that small units in hospitals ALMOST EVERYWHERE will be able to manage neonatal and many cases of adult RDS (like those secondary to bacterial and viral pneumonia) in the setting of the community ICU. The cost savings could be enormous: fewer transfers to tertiary care centers, earlier discharges in community acquired pneumonia, shorter hospital stays.
Of course, there are other advantages which are not so immediately obvious; more revenue for smaller institutions able to manage patients they would have transferred, neonates and sick adults remaining close to their family in their own community (and associated cost and hardship savings), and perhaps most importantly, the ability of third-world and other more medically resource restricted countries to rescue patients they would now lose due to the unavailability of ECMO or other costly and technically demanding ECLS technologies. All of this was a pretty clear implication from Gauger's presentation.
John Ender's presentation on the lung-protective effects of LiquiVent was fascinating. PLV has profound anti-inflammatory effects. Some of this is, no doubt, due to the alveolar lavage effects discussed earlier in this review. Since lung exudate/ secretions are not soluble in LiquiVent, they float to the top and can be suctioned off. Indeed, LiquiVent can in principle be recovered from the suction reservoir and recycled as the secretions float atop it and are not only insoluble but immiscible as well. The alveolar lavage and the ability to use lower peak pressures means less mechanical stress on the lung. Lower FiO2s can be used and this means less oxygen toxicity and less absorption atelectasis due to N2 depletion. One thing which *is* soluble in LiquiVent is NO. Need I say more? NO and LiquiVent together? Are we dreaming, or what? Speaking of magic bullets.... ;)
Concomitant with the ability to reduce vent pressure is the sparing of the inflated alveoli from over-distention. This is achieved in large measure by LiquiVent's remarkable ability to radically improve lung compliance in ARDS and neonatal RDS. These points were dramatically illustrated in a slide displayed by Gauger showing the capillary leak index (CLI) (scale 0-1.4) versus conventional and PLV showing pressure and volume as indices with 10 ml/kg of Liquivent being used. In the barotrauma group the CLI was nearly 1.4. In the barotrauma group treated with Liquivent the index was ca. 1/3rd to 1/4th of that. In the volutrauma group, the CLI was ca. 1.0 to 1.2 with conventional PPV and roughly 1/3rd of that in the in PLV group. If I seem vague here it is because I am working from a *sketch* of this slide I made in my notes. Scaling the ogives accurately was impossible, and it was all I could do to just get the X & Y parameters written down (hopefully correctly!). So, take these numbers with a grain of salt. But I think they are reasonably representative. Gauger also reviewed the direct anti-inflammatory effects of Liquivent in vitro:
* decreased release of lysosomal enzymes
* decreased granulocyte adherence to endothelial cells
* decreased phagocytosis
* decreased chemotaxis
* decreased SOD generation
5 minutes of incubation of neutrophils with LiquiVent was profoundly effective at inhibiting endothelial cell injury. Mechanisms of action are not know for sure at this point but what is known is that LiquiVent *must* be in contact with the neutrophils or endothelial cells to exert this effect. One hypothesis put forward with (it seemed to me preference) was that LiquiVent is physically blocking the sites for activation.
A number of other mechanisms were proposed and I made a sketch of the slide showing a neutrophil above a series of capillary endothelial cells. Two mechanisms of inflammation inhibition were know to be operative, and two were yet to be proved, but were possible: Cytokines? lytic enzymes? Leukocyte adhesion -- yes. Production of oxygen reactive species -- yes
Details of work using an animal c5a complement activation model were presented: Cobra venom was administered which produces very rapid (5 min) development of pulmonary edema. 10 cc/kg of LiquiVent was then used to partially liquid ventilate (PLV) the animals. There was marked decrease in neutrophil recruitment, capillary leak, lung water, and myeloperoxidase production. I did not note and do not remember the animal species used in this study.
This was a very satisfying presentation. Indeed, one of the nice things about this conference, and one of the amazing things, was just how far investigations into different aspects of PLV had proceeded and how elegantly these investigations supported each others' conclusions. The clinical investigations (including animal clinical models) showed what appeared to be less inflammation mediated injury. The in vitro studies provided confirmation and insight into at least a couple of the mechanisms. This was encouraging and often is missing from early studies of a new treatment modality. But, obviously, things are not as early in the game with PLV as they seem. The speed with which this technology has moved from the lab to phase I and now Phase III clinical trials is stunning. Other than AZT I can't think of a single recent example.