X-Message-Number: 7762 Date: 26 Feb 97 07:16:05 EST From: Mike Darwin <> Subject: Copy of: Advances in cryopreservation Recently Brian Wowk posted some disclosures about the (FC) cooling approach being developed at 21st Century Medicine. This prompted a number of questions (all good ones) which I had hoped to have time to answer sooner. As it is, we have all been busy doing ICU care on one of our asanguineous (blood washed out) deep hypothermic circulatory arrest dogs. Since we are using a lot of new technology and pharmacology on this run and it has been sometime since we last pumped a survival dog, we have had our hands full! This post will be messy; I have had little sleep since Saturday. I make no apologies. I will try to address the questions asked from memory, and I will also expand a bit on FC cooling and elucidate some problems and pitfalls which will hopefully put things in a clearer perspective. 1) Thomas Donaldson asks how sure we are this will work. The answer is, pretty sure. Others have perfused organs with pure fluorocarbons (i.e., non emulsified) for reasons not related to heat exchange and have been able to achieve good flows and normal histology and acute metabolism following 24, 48, and 72 hours of storage at near 0 degrees C. I can provide the cite if anyone is interested. We have done several dog runs and we have found a number of significant pitfalls. FCs are totally insoluable in water or cryoprotectant-water mixtures. Most are approximately 2X the density of water and most have kinematic viscosities of far less than that of water and dramatically less than that of blood or cryoprotective perfusate (CPA) mixtures. These three factors mean that getting FCs in and evenly distributed requires clever techniques. The best way to visualize the problem is to imagine trying to deal with air in a complex plumbing system. Air bubbles will float to the top of horizontal pipes and form long, linear bubbles which will be broken up and displaced slowly as fluid is pumped through the system. In this case the "air equivalent" is perfusate and it does not all come out of the system easily. On the low pressure venous side you are left with a long, linear bubble of perfusate running the length of the vena cava. If you tip the animal the perfusate will behave just like air would in such a system since it is less dense; it will quickly flow to the highest point. If you are moving the animal into a cold bath and you stop flow of FC and tip the animal during the move the CPA perfusate will displace FC in the cannulae. If the cannulae are in a bath at below the freezing point of the perfusate you will get a frozen plug of perfusate in the cannulae and/or the connecting lines. This little failure mode cost us a $5,000 experiment! Another problem is the complexity of flow through a multi-pathed circuit which is filled with a high viscosity fluid (CPA perfusate) and which is being flushed with a low viscosity fluid (FC). What will tend to happen is that the shortest connecting paths or shunts between the input and output sides of the plumbing array will be flushed out with FC first. In other words, the shortest paths between the arterial and venous elements of the circulatory system will be flushed clear of CPA perfusate and filled with FC. The problem is that once this happens, these shorter, lower resistance pathways become "short circuits" through which most of the FC flow occurs through. The FC in effect takes the path of least resistance. In a mammal the brain, kidneys and liver are high flow and will tend to get most of the flow of FC. But, even within these organs there are short circuits in the capillary bed. Consider this fact: in a normal, healthy person only about 1/3rd of the capillaries are open to flow at any one time. Most of the capillaries are cut off from flow by the precapillary sphincters which open and close in a pattern called vasomotion. Once one area of tissue gets well perfused it closes to flow and another area ges flow. Then, the previously perfused area opens up again when it needs more nutrients and oxygen. This vasomotion or autoregulation of capillary flow is mediated by induceable nitric oxide production (and other factors). In multisystem organ failure normal vasomotion is disrupted and autoregulation is lost. Cardiac output goes from a normal of 4.5 liters per minute (LPM) to 30 or 40 LPM! This is called hyperdynamic shock and is a devastating medical problem which has occupied a great deal of my research time and attention over the past few years. Most cryonics patient experience loss of normal vasomotion during the agonal period in slow death. The importance of understanding vasomotion is to realize that much of the capillary bed is _not_ being perfused at any one time. Further, low viscosity FCs will find the shortest pathways and preferentially flow at tremendous rates through these pathways. To open up more of the capillary bed would require increasing the pressure (and flow). In small models we've used just this approach. For larger animals the FC viscosity must be carefully modulated during washout of CPA and low viscosity mixtures (with lower pour points) must be ramped in. However, even with this approach we see "perfusate embolization" during FC perfusion up until the point that the CPA perfusate solidifies (by freezing or vitrifying). Perfusate embolization consists of little streams or bubbles of perfusate entering the main venous FC outflow as more and more of the capillary bed is opened up. This presents a tremendous engineering problem because if this water-CPA perfusate is not _completely_ removed from the venous flow before it is pumped back into the arterial circulation it will act exactly like an air embolism or a blood clot. I have solved this problem with a series of separators, traps and filters but it has been costly and difficult. Seeing this phenomenon has helped to explain why when we wash blood out of dogs or people we see a slow accumulation of red cells in the formerly RBC free perfusate once we close the circuit and start recirculating. We have always wondered how we could wash out a human with 20 liters of perfusate (4X blood volume!) and still see the RBC count rise once we closed the circuit. Now we know why: there is a lot of capillary bed which is still blood filled and the lower viscosity cell free perfusate is "channeling" through the paths of least resistance. This also explain why we see a steady stream of RBCs emerge from the patient as cryoprotective perfusion with glycerol proceeds: as the perfusate viscosity rises and the perfusion pressure rises we open up more and more formerly un-perfused capillary bed. We've done several large animal runs with FC perfusion to and from deep subzero temperatures and we're reasonable confident it is workable. It is certainly workable for heads (brains) or other single organs. We are, however, a LONG way from having all the bugs worked out for whole animals such that revival would be possible. This will remain a problem until we can access _most or all_ the capillary beds with FC and displace all or nearly all the cryoprotective perfusate. What we DO have now is a technique that will allow fairly even and controlled rate cooling of cryopatients. We do perfuse most of the vasculature with FC and the ultrastructure looks much better in dogs so treated following cryopreservation. Small animal runs are easier and less costly and we will probably concentrate on these for the foreseeable future because of: 2) Costs. The FC mixture we are using is astronomically expensive. The FC for one dog alone is nearly $3,000. Yes, we can recover some of it and reuse it. But it is still very expensive stuff. This expense has lead us to develop an alternative to the traditional immersion bath which has been used to cool cryopatients for years. Instead of using a tank full of astronomically costly FC the patient will probably be placed on a conductive metal lattice-work elevated above the bottom of a tank and a continuous, high flow spray of FC will be directed over the the surface of the patient. Why use FC for external cooling at all, you may ask? Because some areas may not perfuse well with FC due to clotting, injury, short-circuiting, etc., and we will be relying on surface cooling. Secondly, the FCs we are using have viscosities of 0.5 cs. (about half that of water) at room temperature and the viscosity climbs only to 2-3 cs. at -90 C! FCs also have very low surface tensions and excellent spreading coefficients which means that they will flow through ANY hole or tear in the system. A needle stick becomes a geyser of FC! These agents are used as leak-check agents in industrial applications for just this reason. If you will recall a recent technical brief from me on the use of Super Glue to achieve "hemostasis" in cryopatients you should now understand why this became so critical an issue: the smallest incision will _massively_ leak FC. In patients with median sternotomies the FC will leak out very rapidly and will flow into the tank holding the patient. This leakage will then be recycled through a "clean-up loop" which will remove all particles and fluid (CPA, ice, solid debris, etc). The same FC used to cool the patient externally will be tapped to provide intravascular cooling as well. Which brings us to: 3) Thomas' question about whether FC will wash out cryoprotective and Jan Cotzee's question about FC's penetrability into cells or their components: FCs are best thought of as liquid teflon. Almost nothing will dissolve in them with the exception of some gases (oxygen, CO2 and nitric oxide being the important ones to us). Virtually all cryoprotectants (old and new) are insoluble in fully perfluoronated chemicals. So, cryoprotectant is perfused first and then this is followed by FC perfusion which displaces the perfusate from the vasculature, but of course does not displace the cryoprotectant or water from the tissues. Jan Cotzee's question about FC solubility in biological systems can be most simply answered by stating that FCs do not cross cell membranes or capillary membranes. They thus remain exclusively in the circulatory system. However, and this is an important _however_, fentogram quantities of FCs appear to dissolve into cell membranes that are in contact with FC. For once, we get a lucky break: in living animals this appears to be the mechanism by which FCs radically down-regulate ICAM activity; an important mediator of the immune-inflammatory response. What this means in practice is that in the living mammal FCs greatly reduce adhesion and de-granulation (release of toxic hypohalous acids like chlorine bleach) of white blood cells. They do not inhibit white cell chemotaxis, but do greatly inhibit WBC adhesion and release of destructive chemicals. This is of tremendous benefit to the lung during liquid ventilation with FCs since the WBCs are the primary mediators of pulmonary injury in shock, multisystem organ failure, pneumonia, etc. 4) There were questions about FC toxicity. We have ventilated _many_ dogs with FCs. Some have been ventilated for days with no ill effects. Right now of the 8 or nine dogs in our colony only 1 has not had FC ventilation. Aside from the dog now in the ICU who has other problems (related to being at 0.6 degrees C with no flow for hours ;-)), all are fine with no residual effects from FC toxicity. Perhaps more to the point, the FCs used for liquid ventilation and the others used for intravascular subzero cooling have high vapor pressures and high spreading coefficients which means that they get all over the place. Thus, the staff of 21CM has been breathing FCs almost continually for years. For those of you concerned about the ozone layer; the FCs we use are not contributors to ozone depletion; it is the chlorinated and brominated fluorochemicals that do this. 5) What about the Prometheus Project? Frankly, it is needed more than ever. The FC mix we'd most like to usefor vitrification is simply not available to us and will not be until after 2003. Costs for implementing FC perfusion in a reversible fashion will be well over 21CM's anticipated ability to bear the costs. In fact, the major reason FC perfusion is not on-line for humans now is the incredibly demanding technology required to deliver subzero fluorochemical perfusion and the high cost of the materials. BioPreservation is making incremental purchases of the system components, but the pace is glacially slow (pun intended). We have bough the heat exchanger and the heat exchanger cabinet, but are a LONG way from a working system for humans. As it is, I have enough FC to allow me to perfuse one human under poorly controlled conditions (not good enough to allow for vitrification). 6) Why do we want to FC cool people now, even though we can't yet vitrify them? The answer is that with the current immersion method of cooling used to freeze people now, freezing takes place at relatively high subzero temperatures and this results in a large increase in the concentration of cryoprotectant agent; ice freezes out as pure water. Right now, that means nearly a doubling of the concentration of cryoprotectant that patients are exposed to. The reason thin tissues (skin, corneas) or cells in culture can survive freezing is that a cooling rate of between 0.5 and 1.0 degrees C per minute can be achieved. _This means that once freezing occurs the cells are exposed to hyper-concentrated CPA at high temperatures only briefly._ When we freeze humans this NOT the case. Freezing occurs and then the patient sits exposed to massive concentrations of agent for _hours_ during subsequent cooling to -90 degrees C. Exposure to these concentrations of agents over a time course of many hours appears to be dissolving cell membranes; thus I am no longer at all sanguine that people frozen with today's techniques are coming back as who they were. Somebody may come back, but the question is, WHO? Mary may have a cloned little lamb who's fleece is white as snow, but the question is does Mary the_ lamb_ still know? It is thus critically important that we start achieving cooling rates in humans that are consistent with cellular survival (membrane integrity) during cryopreservation. This is a staggeringly complex and difficult task and it has been rough sledding going it alone; doing this is BPI's job, not 21CM's and BPI doesn't have that kind of money. (If you want to know where every spare nickel I have has been going, now you know). 7) I believe Prometheus quite understandably underestimated the cost to to deliver reversible brain cryopreservation rather than overestimated it. No one should feel bad about this if this turns out to be the case. We have dozens of cryoprotective compounds to evaluate and many, many more times that number of CPA mixtures. This will take time and money. The most promising agent we've found so far costs nearly $2.00 per gram and is no longer being manufactured. Tooling to make this molecule and related ones of interest will be very costly. We already have two organic chemists on retainer, having pretty much hit the end of the road with most of the big and small chemical houses including Sigma-Fluka-Aldrich, 3M, BASF, Dow, and just about everybody else you can think of. Finally, I want to point out that moderating reperfusion injury and dealing with the intracellular cascade of events which occurs following stress to cells is another area we are working on and which I believe we need to continue to work on. Greg Fahy cannot recover _any_ kidneys unless he moderates reperfusion injury with aspirin, prostaglandin and other drugs upon blood reperfusion. His technology for doing this is dated and crude compared to what we now have available. I believe that controlling the intracellular PARS cascade and moderating inducible nitric oxide synthase and other mediators of injury will be critical to achieving suspended animation. I _know_ that this work will pay big dividends in the treatment of dying cryonics patients by allowing us to stop edema and reperfusion injury and allow us to better perfuse people dying now. These technologies will take time and money. Lots of BOTH. Prometheus was never more needed. I guess that just about covers it. Mike Darwin Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=7762