Newsgroups: sci.cryonics Path: spies!sgiblab!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!cbnewse!cbnewsd!att-out!cbnewsl!kqb From: kqb@cbnewsl.cb.att.com (kevin.q.brown) Subject: Freezing Damage - Part 1 Organization: AT&T Bell Laboratories Date: Fri, 4 Dec 1992 17:55:23 GMT Message-ID: <1992Dec4.175523.9470@cbnewsl.cb.att.com> References: <1992Dec1.204713.17920@cbnewsl.cb.att.com> X-Crossposted-To: cryonics mailing list Lines: 321 ------------------------ Forwarded Message ------------------------ > Date: 03 Dec 92 06:49:03 EST > From: Paul Wakfer <70023.3041@CompuServe.COM> <- Mike Darwin > Message-Subject: CRYONICS: Freezing Damage (Darwin) Part 1 Note: This posting is from Mike Darwin There have been several requests for information about the kind of damage done during cryonic suspension. In particular, there have been requests for detailed, objective studies. As a result of this interest I have decided to post a research paper which is now in the (hopefully) final stages of preparation for publication. However, there is a serious shortcoming to this posting, namely that of necessity there can be no accompanying light or electron micrographs. In this case this is a serious handicap, although for many readers the micrographs would mean little. In any event, it is to be hoped that this paper can be prepared for more formal publication within 6 to 12 months. Anyone who wishes to assist me in this capacity should feel free to do so (the EM's and light micrographs need to be captioned and laid out -- a formidable task). Many caveats about the validity of this work are contained in its closing paragraphs. However, I would like to add the following: The presence of pericapillary ice holes has been verified by freeze substitution work done by Dr. Gregory M. Fahy of the Red Cross Organ preservation lab. Similarly Dr. Fahy's freeze-substitution work has documented the presence of massive ice crystals which comprise about 60% of the tissue volume. I gather that Dr. Fahy feels somewhat more optimistic about preservation of neuronal connectivity than I do, but one thing we are both agreed on: our work clearly demonstrates serious histogical disruption with tears or fractures in the brain tissue appearing at approximately 3 to 5 micron intervals. I think it is also fair to say that anyone, layman or neurophysiologist who looks at either the pictures in the study by Darwin, et al., or the pictures generated by Fahy of freeze- substituted brains (showing massive histological disruption by ice) will be given pause for thought about the workability of cryonics. I have great admiration for Dr. Merkle and his work. But I would also point out that in the theoretical domain where there is a will there is almost always a way. Alas, the real world is somewhat harsher. We all want and need to believe very desperately that cryoinjury can and will be reversed. However, there is no direct evidence of this. The kind of damage my colleagues and I observed in the study below is carefully "qualified," to make it good science. However, in this preamble I am a bit freer and I can say that I believe that the damage is at least as bad as we saw -- it would be hard to imagine it being any worse short of calling it a tissue homogenate -- something you get when you run a brain through a blender. It is my gut feel that post-thaw artifactual stirring was not the main reason things look bad. I think things look bad because they *are* bad. Does that mean patients frozen with today's techniques will not be revived? I do not know. Does that mean we should *not* continue to freeze people? No. What it does mean is that we need to do some serious work to improve the situation. The kind of damage we observed and are observing is a consequence of ice formation. At a minimum we can do a great deal to reduce or eliminate ice formation. A major, comprehensive research proposal is under development at this time and should be ready for submission to the cryonics community by late Spring or early Summer. In the meantime my colleagues and I at Biopreservation and Cryovita are working hard to make further improvements on hypothermic brain presevation which will allow us to conduct the necessary cryopreservation work with greater ease. Also in order is a word about Jerry Leaf, who entered suspension over a year ago. This work was completed in the mid 1980's and this paper was completed in draft form circa 1988. Jerry read and commented on the draft shortly after it was written. Most, but not all of his suggested changes were incorporated. The final two paragraphs of summary were written after his suspension. I feel comfortable that Jerry would want this paper published, warts and all. The work that underpins it took a great deal of his time and effort. Indeed, while none of us knew it at the time, this work comprised a major block of Jerry's cryonics-science productive life. In drawing the few conclusion I draw, I have striven to be as objective as Jerry would have been. I have also submitted this work for review by a prominent cryonicist-cryobiologist whom I know Jerry respected greatly. I have incorporated all of this cryobiologists substantive revisions. Finally, to Jerry: may you someday have the pleasure of reading these words and proving us both wrong about the prospects for recovery. Jerry, I miss you more than words can tell. THE EFFECTS OF CRYOPRESERVATION ON THE CAT by Michael Darwin, Jerry Leaf, Hugh L. Hixon I. Introduction II. Materials and Methods III. Effects of Glycerolization IV. Gross Effects of Cooling to and Rewarming From -196*C V. Effects of Cryopreservation on Histology of Selected Tissues VI. Effects of Cryopreservation on Ultrastructure of Selected Tissues VII. Summary and Discussion I. INTRODUCTION The immediate goal of cryonic suspension is to use current cryobiological techniques to preserve the brain structures which encode personal identity adequately enough to allow for resuscitation or reconstruction of the individual should molecular nanotechnology be realized (1,2). Aside from two previous isolated efforts (3,4) there has been virtually no systematic effort to examine the fidelity of histological, ultrastructural, or even gross structural preservation of the brain following cryopreservation in either an animal or human model. While there is a substantial amount of indirect and fragmentary evidence in the cryobiological literature documenting varying degrees of structural preservation in a wide range of mammalian tissues (5,6,7), there is little data of direct relevance to cryonics. In particular, the focus of contemporary cryobiology has been on developing cryopreservation techniques for currently transplantable organs, and this has necessarily excluded extensive cryobiological investigation of the brain, the organ of critical importance to human identity and mentation. The principal objective of this pilot study was to survey the effects of glycerolization, freezing to liquid nitrogen temperature, and rewarming on the physiology, gross structure, histology, and ultrastructure of both the ischemic and non-ischemic adult cats using a preparation protocol similar to the one then in use on human cryonic suspension patients. The non-ischemic group was given the designation Feline Glycerol Perfusion (FGP) and the ischemic group was referred to as Feline Ischemic Glycerol Perfusion (FIGP). The work described in this paper was carried out over a 19-month period from January, 1982 through July, 1983. The perfusate employed in this study was one which was being used in human cryonic suspension operations at that time, the composition of which is given in Table I. The principal cryoprotectant was glycerol. II. MATERIALS AND METHODS Preperfusion Procedures Nine adult cats weighing between 3.4 and 6.0 kg were used in this study. The animals were divided evenly into a non-ischemic and a 24- hour mixed warm/cold ischemic group. All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Institutes of Health (NIH Publication No. 80-23, revised 1978). Anesthesia in both groups was secured by the intraperitoneal administration of 40 mg/kg of sodium pentobarbital. The animals were then intubated and placed on a pressure-cycled respirator. The EKG was monitored throughout the procedure until cardiac arrest occurred. Rectal and esophageal temperatures were continuously monitored during perfusion using YSI type 401 thermistor probes. Following placement of temperature probes, an IV was established in the medial foreleg vein and a drip of Lactated Ringer's was begun to maintain the patency of the IV and support circulating volume during surgergy. Premedication (prior to perfusion) consisted of the IV administration of 1 mg/kg of metubine iodide to inhibit shivering during external and extracorporeal cooling and 420 IU/kg sodium heparin as an anticoagulent. Two 0.77 mm I.D. Argyle Medicut 15" Sentinel line catheters with Pharmaseal K-69 stopcocks attached to the luer fittings of the catheters were placed in the right femoral artery and vein. The catheters were connected to Gould Model P23Db pressure transducers and arterial and venous pressures were monitored throughout the course of perfusion. Surgical Protocol Following placement of the monitoring catheters, the animals were transferred to a tub of crushed ice and positioned for surgery. The chest was shaved and a median sternotomy was performed. The aortic root was cleared of fat and a purse-string suture was placed, through which a 14-gauge angiocath was introduced. The angiocath, which served as the arterial perfusion cannula, was snared in place, connected to the extracorporeal circuit and cleared of air. The pericardium was opened and tented to expose the right atrium. A purse-string suture was placed in the apex of the right atrium and a USCI type 1967 16 fr. venous cannula was introduced and snared in place. Backties were used on both the arterial and venous cannulae to secure them and prevent accidental dislodgment during the course of perfusion. Placement of cannulae is shown in Figure 2. Extracorporeal Circuit The extracorporeal circuit (Figure 3) was of composed of 1/4" and 3/8" medical grade polyvinyl chloride tubing. The circuit consisted of two sections: a recirculating loop to which the animal was connected and a glycerol addition system. The recirculating system consisted of a 10 liter polyethylene reservoir positioned atop a magnetic stirrer, an arterial (recirculating) roller pump, an Erika HPF-200 hemodialyzer which was used as a hollow fiber oxygenator (8) (or alternatively, a Sci-Med Kolobow membrane oxygenator), a Travenol Miniprime pediatric heat exchanger, and a 40-micron Pall LP 1440 pediatric blood filter. The recirculating reservoir was continuously stirred with a 2" teflon-coated magnetic stir bar driven by a Corning PC 353 magnetic stirrer. Temperature was continuously monitored in the arterial line approximately six inches from the arterial cannula using a Sarns in-line thermistor temperature probe and YSI 42SL remote sensing thermometer. Glycerol concentrate was continuously added to the the recirculating system using a Drake-Willock hemodialysis pump. Storage and Reuse of the Extracorporeal Circuit After use the circuit was flushed extensively with filtered tap and distilled water, and then flushed and filled with 3% formaldehyde in distilled water to prevent bacterial overgrowth. Prior to use the circuit was again thoroughly flushed with filtered tap water, and then with filtered distilled water (including both blood and gas sides of the hollow fiber dialyzer; Kolobow oxygenators were not re-used). At the end of the distilled water flush, a test for the presence of residual formaldehyde was performed using Schiff's Reagent. Prior to loading of the perfusate, the circuit was rinsed with 10 liters of clinical grade normal saline to remove any particulates and prevent osmotic dilution of the base perfusate. Pall filters and arterial cannula were not re-used. The circuit was replaced after a maximum of three uses. Preparation of Control Animals Fixative Perfused Two control animals were prepared as per the above. However, the animals were subjected to fixation after induction of anesthesia and placement of cannulae. Fixation was achieved by first perfusing the animals with 500 cc of bicarbonate-buffered Lactated Ringer's containing 50 g/l hydroxyethyl starch (HES) with an average molecular weight of 400,000 to 500,000 supplied by McGaw Pharmaceuticals of Irvine, Ca (pH adjusted to 7.4) to displace blood and facilitate good distribution of fixative, followed immediately by perfusion of 1 liter of modified Karnovsky's fixative (Composition given in Table I). Buffered Ringers-HES perfusate and Karnovsky's solution were filtered through 0.2 micron filters and delivered with the same extracorporeal circuit described above. Immediately following fixative perfusion the animals were dissected and 4-5 mm thick coronal sections of organs were cut, placed in glass screw-cap bottles, and transported, as detailed below, for light or electron microscopy. Straight Frozen Non-ischemic Control One animal was subjected to straight freezing (i.e., not treated with cryoprotectant). Following induction of anesthesia and intubation the animal was supported on a respirator while being externally cooled in a crushed ice-water bath. When the EKG documented profound bradycardia at 26*C, the animal was disconnected from the respirator, placed in a plastic bag, submerged in an isopropanol cooling bath at -10*C, and chilled to dry ice and liquid nitrogen temperature per the same protocol used for the other two experimental groups as described below. Preparation of FGP Animals Following placement of cannulae, FGP animals were subjected to total body washout (TBW) by open-circuit perfusion of 500 cc of glycerol-free perfusate. The extracorporeal circuit was then closed and constant-rate addition of glycerol-containing perfusate was begun. Cryoprotective perfusion continued until the target concentration of glycerol was reached or the supply of glycerol-concentrate perfusate was exhausted. Preparation of FIGP Animals In the FIGP animals, respirator support was discontinued following anasthesia and administration of Metubine. The endotracheal tube was clamped and the ischemic episode was considered to have begun when cardiac arrest was documented by absent EKG. After the start of the ischemic episode the animals were allowed to remain on the operating table at room temperature ( 22*C to 25*C) for a 30 minute period to simulate the typical interval between pronouncement of legal death in a clinical environment and the start of external cooling at that time. During the 30 minute normothermic ischemic interval the femoral cut-down was performed and monitoring lines were placed in the right femoral artery and vein as per the FGP animals. Prior to placement, the monitoring catheters were irrigated with normal saline, and following placement the catheters were filled with 1000 unit/cc of sodium heparin to guard against clot obstruction of the catheter during the post-mortem ischemic period. After the 30 minute normothermic ischemic period the animals were placed in a 1-mil polyethylene bag, transferred to an insulated container in which a bed of crushed ice had been laid down, and covered over with ice. A typical cooling curve for a FIGP animal is presented in Figure 1. FIGP animals were stored on ice in this fashion for a period of 24 hours, after which time they were removed from the container and prepared for perfusion using the surgical and perfusion protocol described above. Perfusate The perfusate was an intracellular formulation which employed sodium glycerophosphate as the impermeant species and hydroxyethyl starch (HES)(av. MW 400,000 - 500,000) as the colloid. The composition of the base perfusate is given in Table I. The pH of the perfusate was adjusted to 7.6 with potassium hydroxide. A pH above 7.7, which would have been "appropriate" to the degree of hypothermia experienced during cryoprotective perfusion (9), was not achievable with this mixture owing to problems with complexing of magnesium and calcium with the phosphate buffer, resulting in an insoluble precipitate. Perfusate components were reagent or USP grade and were dissolved in USP grade water for injection. Perfusate was prefiltered through a Whatman GFB glass filter (a necessary step to remove precipitate) and then passed through a Pall 0.2 micron filter prior to loading into the extracorporeal circuit. Perfusion Perfusion of both groups of animals was begun by carrying out a total body washout (TBW) with the base perfusate in the absence of any cryoprotective agent. In the FGP group washout was achieved within 2 - 3 minutes of the start of open circuit asanguineous perfusion at a flow rate of 160 to 200 cc/min and an average perfusion pressure of 40 mmHg. TBW in the FGP group was considered complete when the hematocrit was unreadable and the venous effluent was clear. This typically was achieved after perfusion of 500 cc of perfusate. Complete blood washout in the FIGP group was virtually impossible to achieve (see "Results" below). A decision was made prior to the start of this study (based on previous clinical experience with ischemic human cryonic suspension patients) not to allow the arterial pressure to exceed 60 mmHg for any significant period of time. Consequently, peak flow rates obtained during both total body washout and subsequent glycerol perfusion in the FIGP group were in the range of 50-60 cc/min at a mean arterial pressure of 50 mmHg. Due to the presence of massive intravascular clotting in the FIGP animals it was necessary to delay placement of the atrial (venous) cannula (lest the drainage holes become plugged with clots) until the large clots present in the right heart and the superior and inferior vena cava had been expressed through the atriotomy. The chest was kept relatively clear of fluid/clots by active suction during this interval. Removal of large clots and reasonable clearing of the effluent was usually achieved in the FIGP group after 15 minutes of open circuit asanguineous perfusion, following which the circuit was closed and the introduction of glycerol was begun. The arterial pO2 of animals in both the FGP and FIGP groups was kept between 600 mmHg and 760 mmHg throughout TBW and subsequent glycerol perfusion. Arterial pH in the FGP animals was between 7.1 and 7.7 and was largely a function of the degree of diligence with which addition of buffer was pursued. Arterial pH in the FIGP group was 6.5 to 7.3. Two of the FIGP animals were not subjected to active buffering during perfusion and as a consequence recovery of pH to more normal values from the acidosis of ischemia (starting pH for FIGP animals was typically 6.5 to 6.6) was not as pronounced. Introduction of glycerol was by constant rate addition of base perfusate formulation made up with 6M glycerol to a recirculating reservoir containing 3 liters of glycerol-free base perfusate. The target terminal tissue glycerol concentration was 3M and the target time course for introduction was 2 hours. The volume of 6M glycerol concentrate required to reach a terminal concentration in the recirculating system (and thus presumably in the animal) was calculated as follows: Vp Mc = --------- Mp Vc + Vp where Mc = Molarity of glycerol in animal and circuit. Mp = Molarity of glycerol concentrate. Vc = Volume of circuit and exchangeable volume of animal.* Vp = Volume of perfusate added. * Assumes an exchangeable water volume of 60% of the preperfusion weight of the animal. Glycerolization of the FGP animals was carried out at 10*C to 12*C. Initial perfusion of FIGP animals was at 4*C to 5*C with warming (facilitated by TBW with warmer perfusate and removal of surface ice packs) to 10*-12*C for cryoprotectant introduction. The lower TBW temperature of the FIGP animals was a consequence of the animals having been refrigerated on ice for the 24 hours preceding perfusion. Following termination of the cryoprotective ramp, the animals were removed from bypass, the aortic cannula was left in place to facilitate prompt reperfusion upon rewarming, and the venous cannula was removed and the right atrium closed. The chest wound was loosely closed using surgical staples. Concurrent with closure of the chest wound, a burrhole craniotomy 3 to 5 mm in diameter was made in the right parietal bone of all animals using a high speed Dremel "hobby" drill. The purpose of the burrhole was to allow for post-perfusion evaluation of cerebral volume, assess the degree of blood washout in the ischemic animals and facilitate rapid expansion of the burrhole on rewarming to allow for the visual evaluation of post-thaw reperfusion (using dye). The rectal thermistor probe used to monitor core temperature during perfusion was replaced by a copper/constantan thermocouple at the conclusion of perfusion for monitoring of the core temperature during cooling to -79*C and -196*C. Cooling to -79*C Cooling to -79*C was carried out by placing the animals within two 1 mil polyethylene bags and submerging them in an isopropanol bath which had been precooled to -10*C. Bath temperature was slowly reduced to -79*C by the periodic addition of dry ice. A typical cooling curve obtained in this fashion is shown in Figure 5. Cooling was at a rate of approximately 4*C per hour. Cooling to and Storage at -196*C Following cooling to -79*C, the plastic bags used to protect the animals from alcohol were removed, the animals were placed inside nylon bags with draw-string closures and were then positioned atop a 6" high aluminum platform in an MVE TA-60 cryogenic dewar to which 2"- 3" of liquid nitrogen had been added. Over a period of approximately 48 hours the liquid nitrogen level was gradually raised until the animal was submerged. A typical cooling curve to liquid nitrogen temperature for animals in this study is shown in Figure 6. Cooling rates to liquid nitrogen temperature were approximately 2*C per hour. After cool-down animals were maintained in liquid nitrogen for a period of 6-8 months until being removed and rewarmed for gross structural, histological, and ultrastructural evaluation. Rewarming The animals in both groups were rewarmed to -2*C to -3*C by removing them from liquid nitrogen and placing them in a precooled box insulated on all sides with a 2" thickness of styrofoam and containing a small quantity of liquid nitrogen. The animals were then allowed to rewarm to approximately -20*C, at which time they were transferred to a mechanical refrigerator at a temperature of 8*C. When the core temperature of the animals had reached -2*C to -3*C the animals were removed to a bed of crushed ice for post-mortem examination and tissue collection for light and electron microscopy. A typical rewarming curve is presented in Figure 7. Modification of Protocol Due To Tissue Fracturing After the completion of the first phase of this study (perfusion and cooling to liquid nitrogen temperature) the authors had the opportunity to evaluate the gross and histological condition of the remains of three human cryonic suspension patients who were removed from cryogenic storage and converted to neuropreservation (thus allowing for post-mortem dissection of the body, excluding the head) (10). The results of this study confirmed previous, preliminary, data indicative of gross fracturing of organs and tissues in animals cooled to and rewarmed from -196*C. These findings led us to abandon our plans to reperfuse the animals in this study with oxygenated, substrate-containing perfusate (to have been followed by fixative perfusion for histological and ultrastructural evaluation) which was to be have been undertaken in an attempt to assess post-thaw viability by evaluation of post-thaw oxygen consumption, glucose uptake, and tissue-specific enzyme release. Rewarming and examination of the first animal in the study confirmed the presence of gross fractures in all organ systems. The scope and severity of these fractures resulted in disruption of the circulatory system, thus precluding any attempt at reperfusion as was originally planned. Preparation of Tissue Samples For Microscopy Fixation Samples of four organs were collected for subsequent histological and ultrastructural examination: brain, heart, liver and kidney. Dissection to obtain the tissue samples was begun as soon as the animals were transferred to crushed ice. The brain was the first organ removed for sampling. The burrhole created at the start of perfusion was rapidly extended to a full craniotomy using rongeurs (Figure 8). The brain was then removed en bloc to a shallow pan containing iced, modified Karnovsky's fixative containing 25% w/v glycerol (see Table I for composition) sufficient to cover it. Slicing of the brain into 5 mm thick sections was carried out with the brain submerged in fixative in this manner. At the conclusion of slicing a 1 mm section of tissue was excised from the visual cortex and fixed in a separate container for electron microscopy. During final sample preparation for electron microscopy care was taken to avoid the cut edgdes of the tissue block in preparing the Epon embedded sections. The sliced brain was then placed in 350 ml of Karnovsky's containing 25%w/v glycerol in a special stirring apparatus which is illustrated in Figure 9. This fixation/deglycerolization apparatus consisted of two plastic containers nested inside of each other atop a magnetic stirrer. The inner container was perforated with numerous 3 mm holes and acted to protect the brain slices from the stir bar which continuously circulated the fixative over the slices. The stirring reduced the likelihood of delayed or poor fixation due to overlap of slices or stable zones of tissue water stratification. (The latter was a very real possibility owing to the high viscosity of the 25%w/v glycerol-containing Karnovsky's.) Deglycerolization of Samples To avoid osmotic shock all tissue samples were initially immersed in Karnovsky's containing 25%w/v glycerol at room temperature and were subsequently deglycerolized prior to staining and embedding by stepwise incubation in Karnovsky's containing decreasing concentrations of glycerol (see Figure 10 for deglycerolization protocol). To prepare tissue sections from heart, liver, and kidney for microscopy, the organs were first removed en bloc to a beaker containing an amount of ice-cold fixative containing 25% w/v glycerol sufficient to cover the organ. The organ was then removed to a room temperature work surface at where 0.5 mm sections were made with a Stadie-Riggs microtome. The microtome and blade were pre-wetted with fixative, and cut sections were irrigated from the microtome chamber into a beaker containing 200 ml of room-temperature fixative using a plastic squeeze-type laboratory rinse bottle containing fixative solution. Sections were deglycerolized using the same procedure previously detailed for the other slices. Osmication and Further Processing At the conclusion of deglycerolization of the specimens all tissues were separated into two groups; tissues to be evaluated by light microscopy, and those to be examined with transmission electron microscopy. Tissues for light microscopy were shipped in glycerol- free modified Karnovsky's solution to American Histolabs, Inc. in Rockville, MD for paraffin embedding, sectioning, mounting, and staining. Tissues for electron microscopy were transported to the facilities of the University of California at San Diego in glycerol- free Karnovsky's at 1* to 2*C for osmication, Epon embedding, and EM preparation of micrographs by Dr. Paul Farnsworth. Due to concerns about the osmication and preparation of the material processed for electron microscopy by Farnsworth, tissues from the same animals were also submitted for electron microscopy to Electronucleonics of Silver Spring, Maryland. ***Electronucleonics results are not covered here since another investigator has yet to provide the necessary information and we do not have access to the pictures. III. EFFECTS OF GLYCEROLIZATION Perfusion of FGP Animals Blood washout was rapid and complete in the FGP animals and vascular resistance decreased markedly following blood washout. Vascular resistance increased steadily as the glycerol concentration increased, probably as a result of the increasing viscosity of the perfusate. Within approximately 5 minutes of the beginning of the cryoprotective ramp, bilateral ocular flaccidity was noted in the FGP animals. As the perfusion proceeded, ocular flaccidity progressed until the eyes had lost approximately 30% to 50% of their volume. Gross examination of the eyes revealed that initial water loss was primarily from the aqueous humor, with more significant losses from the posterior chamber of the eyes apparently not occurring until later in the course of perfusion. Within 15 minutes of the start of glycerolization the corneal surface became dimpled and irregular and the eyes had developed a "caved-in" appearance. Dehydration was also apparent in the skin and skeletal muscles and was evidenced by a marked decrease in limb girth, profound muscular rigidity, cutaneous wrinkling (Figure 11), and a "waxy- leathery" appearance and texture to both cut skin and skeletal muscle. Tissue water evaluations conducted on ileum, kidney, liver, lung, and skeletal muscle confirmed and extended the gross observations. Preliminary observation suggest that water loss was in the range of 30% to 40% in most tissues. As can be seen in Table III, total body water losses attributable to dehydration, while typically not as profound, were still in the range of 18% to 34%. The gross appearance of the heart suggested a similar degree of dehydration, as evidenced by modest shrinkage and the development of a "pebbly" surface texture and a somewhat translucent or "waxy" appearance. Examination of the cerebral hemispheres through the burr hole (Figure 12) revealed an estimated 30% to 50% reduction in cerebral volume, presumably as a result of osmotic dehydration secondary to glycerolization. The cortices also had the "waxy" amber appearance previously observed as characteristic of glycerolized brains. The gross appearance of the kidneys, spleen, mesenteric and subcutaneous fat, pancreas, and reproductive organs (where present) were unremarkable. The ileum and mesentery appeared somewhat dehydrated, but did not exhibit the waxy appearance that was characteristic of muscle, skin, and brain. Oxygen consumption (determined by measuring the arterial/venois difference) throughout perfusion was fairly constant and did not appear to be significantly impacted by glycerolization, as can be seen Figure 12. Perfusion of FIGP Animals As previously noted, the ischemic animals had far lower flowrates at the same perfusion pressure as FGP animals and demonstrated incomplete blood washout. Intravascular clotting was serious a barrier to adequate perfusion. Post-thaw dissection demonstrated multiple infarcted areas in virtually all organ systems; areas where blood washout and glycerolization were incomplete or absent. In contrast to the even color and texture changes observed in the FGP animals, the skin of the FIGP animals developed multiple, patchy, nonperfused areas which were clearly outlined by surrounding, dehydrated, amber-colored glycerolized areas. External and internal examination of the brain and spinal cord revealed surprisingly good blood washout of the central nervous system. While grossly visible infarcted areas were noted, these were relatively few and were generally no larger than 2 mm to 3 mm in diameter. With few exceptions, the pial vessels were free of blood and appeared empty of gross emboli. One striking difference which was consistently observed in FIGP animals was a far less profound reduction in brain volume during glycerolization (Figure 13). This may have been due to a number of factors: lower flow rates, higher perfusion pressures, and the increased capillary permeability and perhaps increased cellular permeability to glycerol. Whereas edema was virtually never a problem during glycerolization of FGP animals, edema was universal in the FIGP animals after as little as 30 minutes of perfusion. In the central nervous system this edema was evidenced by a "rebound" from initial cerebral shrinkage to frank cerebral edema, with the cortices, restrained by the dura, often abutting or slightly projecting into the burrhole. Marked edema of the nictating membranes, the lung, the intestines, and the pancreas was also a uniform finding at the conclusion of cryoprotective perfusion. The development of edema in the central nervous system sometimes closely paralleled the beginning of "rebound" of ocular volume and the development of ocular turgor and frank ocular edema. In contrast to the relatively good blood washout observed in the brain, the kidneys of FIGP animals had a very dark and mottled appearance. While some areas (an estimated 20% of the cortical surface) appeared to be blood-free, most of the organ remained blood- filled throughout perfusion. Smears of vascular fluid made from renal biopsies which were collected at the conclusion of perfusion (for tissue water determinations) revealed the presence of many free and irregularly clumped groups of crenated and normal-appearing red cells, further evidence of the incompleteness of blood washout. Microscopic examination of recirculating perfusate revealed some free, and a few clumped red cells. However, the concentration was low, and the perfusate microhematocrit was unreadable at the termination of perfusion (i.e., less than 1%). The liver of FIGP animals appeared uniformly blood-filled throughout perfusion, and did not exhibit even the partial blood washout evidenced by the kidneys. However, despite the absence of any grossly apparent blood washout, tissue water evaluations in one FIGP animal were indicative of osmotic dehydration and thus of some perfusion. The mesenteric, pancreatic, splanchic, and other small abdominal vessels were largely free of blood by the conclusion of perfusion. However, blood-filled vessels were not uncommon, and examination during perfusion of mesenteric vessels performed with an ophthalmoscope at 20X magnification revealed stasis in many smaller vessels, and irregularly shaped small clots or agglutinated masses of red cells in most of the mesenteric vessels. Nevertheless, despite the presence of massive intravascular clotting, perfusion was possible, and significant amounts of tissue water appear to have been exchanged for glycerol. One immediately apparent difference between the FGP and FIGP animals was the accumulation in the lumen of the ileum of large amounts of perfusate or perfusate ultrafiltrate by the ischemic animals. Within approximately 10 minutes of the start of reperfusion, the ileum of the ischemic animals that had been laparotomized was noticed to be accumulating fluid. By the end of perfusion, the stomach and the small and large bowel had become massively distended with perfusate. Figure 14 shows both FIGP and FGP ileum at the conclusion of glycerol perfusion. As can be clearly seen, the FIGP intestine is markedly distended. Gross examination of the gut wall was indicative of tissue-wall edema as well as intraluminal accumulation of fluid. Often by the end of perfusion, the gut had become so edematous and distended with perfusate that it was impossible to completely close the laparotomy incision. Similarly, gross examination of gastric mucosa revealed severe erosion with the mucosa being very friable and frankly hemorrhagic. Escape of perfusate/stomach contents from the mouth (purging) which occurs during perfusion in ischemically injured human suspension patients did not occur, perhaps due to greater post-mortem competence of the gastroesophageal valve in the cat. Oxygen consumption in the two ischemic cats in which it was measured was dramatically impacted, being only 30% to 50% of control and deteriorating throughout the course of perfusion (Figure 12). IV. GROSS EFFECTS OF COOLING TO AND REWARMING FROM -196*C The most striking change noted upon thawing of the animals was the presence of multiple fractures in all organ systems. As had been previously noted in human cryonic suspension patients, fracturing was most pronounced in delicate, high flow organs which are poorly fiber- reinforced. An exception to this was the large arteries such as the aorta, which were heavily fractured. Fractures were most serious in the brain, spleen, pancreas, and kidney. In these organs fractures would often completely divide or sever the organ into one or more discrete pieces. Tougher, more fiber-reinforced tissues such as myocardium, skeletal muscle, and skin were less affected by fracturing; there were fewer fractures and they were smaller and less frequently penetrated the full thickness of the organ. In both FGP and FIGP animals the brain was particularly affected by fracturing and it was not uncommon to find fractures in the cerebral hemispheres penetrating through to the ventricles as seen in Figure 15, or to find most of both cerebral hemispheres and the mid- brain completely severed from the cerebellum by a fracture (Figure 16). Similarly, the cerebellum was uniformly severed from the medulla at the foramen magnum as were the olfactory lobes, which were usually retained within the olfactory fossa with severing fractures having occurred at about the level of the transverse ridge. The spinal cord was invariably transversely fractured at intervals of 5 mm to 15 mm over its entire length. Bisecting CNS fractures were most often observed to occur transversely rather than longitudinally. In general, roughly cylindrical structures such as arteries, cerebral hemispheres, spinal cord, lungs, and so on are completely severed only by transverse fractures. Longitudinal fractures tend to be shorter in length and shallower in depth, although there were numerous exceptions to this generalization. In ischemic animals the kidney was usually grossly fractured in one or two locations (Figure 17). By contrast, the well-perfused kidneys of the nonischemic FGP group exhibited multiple fractures, as can be seen in Figure 18. A similar pattern was observed in other organ systems as well; the nonischemic animals experienced greater fracturing injury than the ischemic animals, presumably as a result of the higher terminal glycerol concentrations achieved in the nonischemic group. Cannulae and attached stopcocks where they were externalized on the animals were also frequently fractured. In particular, the polyethylene pressure-monitoring catheters were usually fractured into many small pieces. The extensive fracture damage occurring in cannulae, stopcocks, and catheters was almost certainly a result of handling the animals after cooling to deep subzero temperatures, as this kind of fracturing was not observed in these items upon cooling to liquid nitrogen temperature (even at moderate rates). It is also possible that repeated transfer of the animals after cooling to liquid nitrogen temperature may have contributed to fracturing of tissues, although the occurrence of fractures in organs and bulk quantities of water-cryoprotectant solutions in the absence of handling is well documented in the literature (12, 13). There were subtle post-thaw alterations in the appearance of the tissues of all three groups of animals. There was little if any fluid present in the vasculature and yet the tissues exhibited oozing and "drip" (similar to that observed in the muscle of frozen-thawed meat and seafood) when cut. This was most pronounced in the straight- frozen animal. The tissues (especially in the ischemic group) also had a somewhat pulpy texture on handling as contrasted with that of unfrozen, glycerolized tissues (i.e., those handled during pre- freezing sampling for water content). This was most in evidence by the accumulation during the course of dissection of small particles of what appeared to be tissue substance with a starchy appearance and an oily texture on gloves and instruments . This phenomenon was never observed when handling fresh tissue or glycerolized tissue prior to freezing and thawing. There were marked differences in the color of the tissues between the three groups of animals as well. This was most pronounced in the straight-frozen control where the color of almost every organ and tissue examined had undergone change. Typically the color of tissues in the straight-frozen animal was darker, and white or translucent tissues such as the brain or mesentery were discolored with hemoglobin released from lysed red cells. The FGP and FIGP groups did not experience the profound post-thaw changes in tissue color experienced by the straight-frozen controls, although the livers and kidneys of the FIGP animals appeared very dark, even when contrasted with their pre-perfusion color as observed in those animals laparotomized for tissue water evaluation. IV. EFFECTS OF CRYOPRESERVATION ON THE HISTOLOGY OF SELECTED TISSUES Histology was evaluated in two animals each from the FIG and FIGP groups, and in one control animal. Only brain histology was evaluated in the straight-frozen control animal. Liver The histological appearance of the liver in all three groups of animals was one of profound injury. Even in the FGP group the cellular integrity of the liver appeared grossly disrupted. In liver tissue prepared using Yajima stain the sinusoids and spaces of Disse were filled with flocculent debris and it was often difficult or impossible to discern cell membranes. The collagenous supporting structures of the bile canaliculi were in evidence and the nuclei of the hepatocytes appeared to have survived with few alterations evident at the light level, although occasional pyknotic nuclei were noted in the FIGP group. Indeed, the nuclei often appeared to be floating in a sea of amorphous material. Not surprisingly, the density of staining of the cytoplasmic material was noticeably reduced over that of the fixative-perfused control. Few intact capillaries were noted. FGP liver tissue prepared with PAS stain exhibited a similar degree of disruption. However, quite remarkably, the borders of the hepatocytes were defined by a clear margin between glycogen granule containing cytoplasm and non-glycogen containing membrane or other material (membrane debris?) which failed to stain with Yajima stain due to gross disruption or altered chemistry. Kidney PAS stain was used to prepare the control, FGP and FIGP renal tissue for light microscopy. The histological appearance of FGP renal tissue was surprisingly good. The glomeruli and and tubules appeared grossly intact and stain uptake was normal. However, a number of alterations from the appearance of the control were apparent. The capillary tuft of the glomeruli appeared swollen and the normal space between the capillary tuft and Bowman's capsule was absent. There was also marked interstitial edema, and marked cellular edema as evidenced by the obliteration of the tubule lumen by cellular edema. By contrast, the renal cortex of the FIGP animals, when compared to either the control or the FGP group, showed a profound loss of detail, absent intercellular space, and altered staining. The tissue appeared frankly necrotic, with numerous pyknotic nuclei and numerous large vacuoles which peppered the cells. One striking difference between FGP and FIGP renal cortex was that the capillaries, which were largely obliterated in the FGP animals, were consistently spared in the FIGP animals. Indeed, the only extracellular space in evidence in this preparation was the narrowed lumen of the capillaries, grossly reduced in size apparently as a consequence of cellular edema. Both ischemic and nonischemic sections showed occasional evidence of fracturing, with fractures crossing and severing tubule cells and glomeruli. Cardiac Muscle Yajima stain was used to prepare the Control, FGP and FIGP cardiac tissue for light microscopy. The histological appearance of FGP cardiac muscle was grossly normal with one exception; there was increased interstitial space, probably indicative of interstitial edema. The banding pattern was normal and the nuclei were unremarkable. Similarly, FIGP cardiac tissue appeared relatively normal histologically. The principal alterations from control and from the FGP group were the noticeable presence of increased interstitial space and a more "ragged" or rough appearance of the myofibrils where they are silhouetted against interstitial space. Most surprising was the general absence of thaw-rigor in the FGP group and only the occasional presence of rigor in the FIGP group. No microscopic evidence of fracturing was noted in either the FGP or the FIGP groups. Brain Bodian stain was used to prepare the control, FGP, and FIGP brain tissue samples for light microscopy. Three striking changes were apparent in FGP cerebral cortex histology: 1) marked dehydration of both cells and cell nuclei, 2) the presence of tears or cuts at intervals of 10 to 30 microns throughout the tissue on a variable basis (some areas were spared while others were heavily lesioned), and 3) the increased presence (over control) of irregular, empty spaces in the neuropil as well as the occasional presence of large pericapillary spaces. These changes were fairly uniform throughout both the molecular layer and the second layer of the cerebral cortex. Changes in the white matter paralleled those in the cortex with the notable exception that dehydration appeared to be more pronounced. Other than the above changes, both gray and white matter histology appeared remarkably intact, and only careful inspection could distinguish it from control. The neuropil appeared normal (aside from the aforementioned holes and tears) and many long axons could be observed traversing the field. Cell membranes appeared crisp and apart from appearing dehydrated, neuronal architecture appeared comparable to control. Similarly, staining was comparable to that observed in control cerebral cortex. Cell-to-cell connections appeared largely undisrupted. The histological appearance of FIGP brain differed from that of FGP animals in that ischemic changes such as the presence of pyknotic and fractured nuclei were much in evidence and cavities and tears in the neuropil appeared somewhat more frequently. Both FGP and FIGP brains presented occasional evidence of microscopic fractures. V. EFFECTS OF CRYOPRESERVATION ON THE ULTRASTRUCTURE OF SELECTED TISSUES Ultrastructure was evaluated in two animals from the FIG and FIGP groups, and in one control animal. Only brain ultrastructure was evaluated in one straight-frozen control animal. Liver Hepatic ultrastructure was grossly disrupted, with the tissue presenting more as a homogenate than as an organized tissue. While organelle membranes, particularly rough endoplasmic reticulum, nuclear membranes, and mitochondrial membranes were frequently intact, the presence of intact cell membranes was the exception rather than the rule. The sinusoids, bile canaliculi, and capillaries, where these structures were identifiable, were largely filled with debris. The character of this debris ranged from the relatively amorphous granular and flocculent debris observed in the other organ systems of FGP and FIGP animals to relatively organized fragments of cytosol, free organelles (naked nuclei and mitochondria being the most frequently observed), as well as somewhat structured but unidentifiable debris. In areas where discernible hepatocyte membranes were visible, the intracellular contents appeared washed out and depleted of ground substance. Similarly, the spaces of Disse were hard to identify and where identifiable were both collapsed and filled with debris. In the FIGP animals, intact red cells were frequently in evidence as well as sinusoids full of what appeared to be leukocytes and/or leukocyte debris, indicating failed blood washout and probable failed cryoprotective perfusion as well. Mitochondria, where identifiable, rarely had much ground substance and presented only faint evidence of cristae. Nuclei in the livers of both FGP and FIGP animals appeared reasonably well preserved and the double nuclear membrane was frequently (although not universally) intact. Kidney The ultrastructure of FGP renal tissue was intact to a surprising degree. The desmosomes, endoplasmic reticulum, and intracellular organelles, with the exception of the mitochondria appeared normal. Most mitochondria demonstrated marked enlargement, decreased matrix density, disruption of cristae and a few amorphous matrix densities. The nuclei were largely free of margination and clumping of chromatin. The glomeruli appeared intact as did tubule and mesangial cells. The architecture of the brush border and urinary space compared favorably to control with little debris in evidence. While there was little debris in the intercellular spaces, there was extensive debris in the capillary spaces, where it was common to find the capillary completely obliterated and free red cells present. Intact capillaries were occasionally observed in FGP renal tissue. However, this was the exception rather than the rule. By contrast, the capillaries in the FIGP animals were more consistently intact. The narrow lumens of these relatively well preserved capillaries constituted virtually the only extracellular space visible. Also remarkable, given the poor appearance of the tissue at the light level, was the presence of a considerable amount of renal ultrastructure. The microvilli, glomeruli, and the mesangial cells were all present and reasonably intact. However, the urinary space and capillary lumens were filled with flocculent debris. Ultrastructural changes in cell organelles were more pronounced with the nuclei exhibiting clumping of the chromatin, the consistent presence of megamitochondria exhibiting loss of integrity of membranes and many amorphous matrix densities. Heart Cardiac ultrastructure in both FGP and FIGP animals was reasonably well preserved. The sarcoplasmic reticulum, transverse tubules, intercalated discs, and banding of the myofibrils were comparable to that of control. A notable abnormality in both the FGP and FIGP myocardium was the presence of severe interstitial edema, as evidenced by greatly increased interstitial spaces littered with both granular and flocculent debris. There did not appear to be a significant difference in the quantity, character, or location of debris between the FGP and FIGP animals. No significant amount of fibrolysis was noted in either the FGP or the FIGP groups. Also notable was the presence of megamitochondria, with decreased matrix density and disruption of cristae. Occasional mitochondria with normal density were observed in FGP animals. However, this was virtually never the case in the FIGP animals. Myocardial capillaries were grossly intact, with only the infrequent presence of what appeared to be very small areas of focal injury involving separation of the endothelial cell membrane from basement membrane. Small areas of rigor evidenced by the presence of severe contraction bands were sometimes present in the FIGP group but were not noted in the FGP group. FGP Brain At the outset it should be noted that evaluation of the fine ultrastructure of FGP cerebral cortex is complicated by the degree of apparent dehydration of intracellular structures present. Ground substance was markedly increased over control and most, though not all, axons appeared shrunken, electron-dense, and surrounded by a periaxonal space. Intraorganelle structures were frequently difficult to identify as a result of dehydration, with many structures presenting an electron dense but amorphous interior. These effects notwithstanding, the overall architecture of the tissue could be discerned. Intact neuronal, glial, and vascular cell membranes were uniformly present. Such interstitial space as was present consisted of periaxonal shrinkage spaces and irregularly shaped cavities, apparently artifacts of ice formation, of widely varying size, often containing small quantities of organized debris, which peppered the tissues at intervals of 5 to 10 microns. The largest of these cavities appeared to be 20 to 30 microns across and presented the appearance of tears or rips, with the two opposing sides of the gap presenting a rough match. Smaller cavities 1-3 microns in diameter were more frequently present than these relatively large tears. The capillaries appeared reasonably intact, with the lumens containing no or modest amounts of relatively well organized debris. However, the capillaries were frequently surrounded by cavities. These cavities varied in size from a few microns to 10 to 15 microns in diameter with the cavity separating the capillary from surrounding brain cells usually circumscribing from one-third to one-half of the capillary perimeter. Axons usually appeared intact, but shrunken. However, it should be noted that some spaces characteristic of axons and containing myelin debris (but no axon) were also present in most sections of FGP brain examined. Myelinated tracts were often difficult to evaluate due to the degree of dehydration. However, unraveled and disrupted myelin was commonplace, often surrounding an intact-looking axon. Synapses were present in numbers comparable to that seen in the control and were especially well preserved, presenting grossly normal architecture, including clear pre- and post-synaptic densities and the presence of synaptic vesicles. The nuclei were highly condensed (presumably an artifact of dehydration) and sometimes contained unusual gaps or spaces which might have occurred as a result of dehydration from glycerolization, the formation of intranuclear ice, or ultramicroscopic fractures as a result of differential contraction during cooling below Tg. The mitochondria appeared dense and amorphous. FIGP Brain The FIGP brains presented an "exploded" appearance at the ultrastructural level. Virtually every structure appeared swollen and there were large amounts of interstitial space. A uniform but not universal alteration was massive swelling and unraveling of the myelin. Typically there was about a 5-fold increase in the thickness of the myelin sheath, with a corresponding decrease in electron density. Often the individual sheets or "turns" of myelin could be easily discerned, with separating spaces between each layer. The presence or absence of intact axons within this disrupted myelin was highly variable; in some regions the axons appeared well preserved, with neurofibrils and microtubules clearly visible, while in others apparently nothing but debris remained. Mitochondria were uniformly swollen and presented varying degrees of internal structure ranging from easily identifiable cristae to a fine-grained amorphous appearance. In contrast to FGP brains there was virtually no dehydration in evidence in the FIGP brains and intracellular structures and small processes such as neurites, where intact, were easily identified. The nuclei appeared more like those present in the control and did not show the peculiar gaps or cavities present in the FGP group. Small cavities and large gaps peppered the tissue as in the FGP cerebral cortex. These cavities contained considerably more debris than those in the FGP brains and the debris were less structured and frequently appeared flocculent and/or granular in nature. Cell membranes were frequently disrupted and masses of free cytosol were common. Synapses, synaptic vesicles and what appeared to be occasional synaptic debris were noted with a frequency comparable to that of the control. VI. SUMMARY AND DISCUSSION Glycerolization Cryoprotective perfusion of non-ischemically injured animals resulted in profound dehydration. This dehydration was particularly pronounced (in terms of visual appearance) in the brain, eyes, skeletal muscle, and skin. While it can be argued that removal of interstitial and intracellular water may be useful in minimizing mechanical injury during subsequent freezing since less water means less ice, it can also be argued that glycerol is failing to adequately penetrate cells and thus is providing less than optimum cryoprotection. Certainly the profound dehydration documented in these animals (and similarly noted in human patients) is indicative of a failure of cellular equilibration of glycerol, particularly in the brain and skeletal muscle, and in and of itself is probably a significant source of osmotic injury. In an unpublished pilot study we tried to determine if better glycerol equilibration could be facilitated by carrying out cryoprotective perfusion at 18*C. Both the gross effects of dehydration and the measured water losses from tissues (including in the brain, which was determined to be 28% in the single experiment conducted) indicated that glycerolizing at higher temperatures is not the solution to this problem. Clinically it has been known for many years that infusion of significant amounts of glycerol at normal body temperatures, as in the case of inadvertent transfusion of frozen- thawed red cells without deglycerolization, results in rapid death from cerebral dehydration (14). Indeed, glycerol has been used as an osmotic agent to control cerebral edema in the traumatized brain (15). Thus, glycerol would seem to be a poor choice of cryoprotectant, at least in terms of its cellular permeability, for the brain. Clearly, a cryoprotective agent(s) capable of better equilibration with the intracellular space of the brain is needed. In the ischemic animals, the gross effects of dehydration were less obvious or were not seen due to the occurrence of interstitial edema. However, cellular dehydration might not have occurred in these animals, perhaps as a result of increased cell membrane permeability due to ischemic changes such as phospolipase (16) or free radical (17) mediated degradation of cellular and organelle membranes. Certainly the intracellular organelles and axons did not have the dense, collapsed, dehydrated appearance of these structures in the nonischemic animals. This noticeable change in cellular glycerol permeability, the loss of capillary integrity as evidenced by the development of serious interstitial edema in the brain and virtually all other body organs with the exception of the liver (which apparently failed to perfuse significantly), the patchy nature of perfusion due to clotting, and the failure to reach target glycerol concentration as a result of all of these effects is indicative of the profound deleterious impact of ischemia and of the importance of minimizing ischemic time and inhibiting mechanisms of ischemic pathology in human suspension patients if adequate distribution and terminal concentration of cryoprotectant is to be achieved. Histology The histological preservation achieved in brain, kidney, and heart in both ischemic (excluding ischemia-associated alterations to nuclei) and non-ischemic animals was surprisingly good considering the magnitude of the insult. In the case of the FGP brains structural preservation appeared excellent and almost indistinguishable from control, with the exceptions of the presence of an increased number of empty cavities and more light-lucent areas, and the presence of obvious tears at 10 to 20 micron intervals in the neuropil. Similarly, the histological preservation of the renal cortex was surprisingly good in both the FGP and the FIGP animals. The glomeruli were generally intact and this is surprising considering the body of data from renal cryopreservation studies documenting destruction of the glomerulus due to ice formation (18, 19). Perhaps the reason this did not occur in our animals was the very slow rate at which cooling was carried out (4*C/hour) as contrasted with the comparatively rapid rate at which kidneys are cooled during cryopreservation experiments. Such comparatively slow cooling rates may have allowed time for water to migrate out of the glomerulus to other sites during freezing (20), and/or the distortive and disruptive effects of ice formation may have been minimized by the plasticity of these structures at the higher temperatures at which most ice formation and growth occurs. Histological preservation in cardiac tissue in both FGP and FIGP animals was also remarkably good and it was often difficult to distinguish ischemic from non-ischemic tissue without careful observation. Ultrastructure The ultrastructural preservation of the brain was unexpectedly poor in all three groups of animals: ischemic, non-ischemic and straight-frozen. Not unexpectedly, the straight-frozen animal presented the worst ultrastructural appearance. The ischemic animals also suffered extensive ultrastructural disruption. This was somewhat unexpected given the relatively good appearance of brain tissue at the light level; in particular it appeared that membranes were crisp and well preserved that cellular ground substance was of reasonably normal density, and that the overall ground substance density of the neuropil, as well as the preservation of long individual axon fibers and cell-to-cell connections, were largely intact. Unfortunately, the degree of ultrastructural injury observed was in sharp contrast to the apparently good histological preservation. The profound loss of ground substance, gross and widespread loss of membrane integrity, presence of extensive debris, and the widespread destruction of the myelin all underscore, yet again, the critical importance of protection of suspension patients from cerebral ischemia. While the degree of ultrastructural disruption was not as profound in the brains of the FGP animals, it was far from acceptable. The presence of frequent ice holes, tears in the neuropil, and the cellular dehydration and fracturing observed are all indicative of unacceptably poor preservation and point to the urgent need for additional research to ameliorate or eliminate these problems. Given the severity of the ultrastructural disruption observed in the brains of all three groups of animals, it is certainly open to question whether or not sufficient structure is being preserved to allow for resuscitation of cryonic suspension patients treated with similar techniques (and presumably injured comparably) with their memories and personalities intact. Freezing Versus Thawing The especially poor perfusion of the liver in the ischemic animals was unexpected. Additionally, the poor ultrastructural preservation observed in the nonischemic animals is puzzling, especially in light of the apparent good perfusion and amounts of water loss (which were comparable to those experienced by the heart and kidney during glycerolization). The relatively good ultrastructure of the kidney and heart in the FGP and to a lesser extent in the FIGP group stand in sharp contrast to widespread disruption seen in the brain. The reason(s) for this are not clear. However, a possible explanation might be the failure of glycerol to penetrate brain cells and provide adequate cryoprotection. It should be noted that the amount of water lost from the brain during glycerolization, while not directly measured, appeared by gross examination to be roughly comparable to that observed in the heart and kidney, both of which were, by comparison, much better preserved. Some caveats regarding these results should be considered. First of all, examination of the tissues was conducted following thawing. This introduces the possibility of significant "stirring" of damaged structure not only during thawing, but also during sectioning and fixation, since re-perfusion with fixative was not possible owing to disruption of the vasculature by fractures. This is potentially a particularly troubling "artifact" because a major concern is the presence of debris many microns from the likely source of origin (as observed in the liver and brain). When and how this debris was translocated from its point of origin, as well as its character (i.e., how unique are the fragments of debris; can their precise point of origin and orientation be determined?) is of critical importance in determining whether or not repair can be undertaken. If the extensive ultrastructural and molecular-level stirring observed in these animals occurred as a result of diffusion/stirring which took place during, or even after thawing and/or during sectioning and fixation, then the situation is considerably more hopeful than if the damage occurred during the freezing process. It will not be easy to determine how much of the observed disruption is a result of freezing, and how much is a result of thawing and/or post-thaw diffusion-driven processes. Depending upon the degree to which the microvasculature is intact following freezing and thawing it should be possible to eliminate pre-fixation sectioning and handling of the tissue as a source of artifacts by the expedient of not cooling to below the glass transition temperature of the the water-cryoprotectant mixture, thus effectively avoiding fracturing and allowing for fixative reperfusion upon thawing. However, evaluating the degree to which freezing, as opposed to freezing followed by thawing, results in the disruption of, and perhaps more importantly, the translocation of cell structures would not be resolved by this means. Finally, it is especially important to point out that this was a pilot study. During the evaluation of the light and electron microscopy it became apparent that additional control groups were needed to resolve many important questions left unanswered by this work. In particular, post glycerolization/pre-freezing ultrastructure and histology should have been evaluated to separate the effects of glycerolization from the effects of subsequent cryopreservation. Similarly, a group of post-thaw cryopreserved tissues should have been deglycerolized prior to fixation in order to allow for evaluation without the confounding effects of glycerol-induced dehydration. Freeze substitution studies at both the light and EM levels would also be useful in helping to relate the lesions observed (gaps, tears, cavities and so on) to mechanical injury resulting from the presence of ice. Technical Issues Some of our most serious caveats are technical in nature. Brain slicing with a Stadie-Riggs microtome could have obliterated structure in and of itself, particularly if the frozen-thawed brain is structurally weaker than a control brain (as is to be expected) . However, this criticism does not apply to the the electron microscopy since tissue examined by EM was from the center of the slice, away from the cut surface. Summary Evaluation of a cryopreservation protocol which is broadly similar to that being used in human cryonic suspensions today discloses poor ultrastructural preservation of the brain, the target organ of the preservation process. The comparatively good ultrastructural preservation of the heart and kidney indicate that better results are possible and strongly suggest that the preservation protocol currently in use is not optimal for the brain and results in unacceptable levels of ultrastructural disruption. There is an urgent need for additional research to address this problem. The impact of prolonged ischemia on tissue histology, ultrastructure, and perfusion was profound and underscores the need to protect suspension patients from ischemia. TABLE I. Composition Of Modified Karnovsky's Solution Component g/l Paraformaldehyde 40 Glutaraldehyde 20 Sodium Chloride 0.2 Sodium Phosphate 1.42 Calcium Chloride 2.0 mM pH adjusted to 7.4 with sodium hydroxide. _________________________________________ TABLE II Perfusate Composition Component mM Potassium Chloride 2.8 Dibasic Potassium Phosphate 5.9 Sodium Bicarbonate 10.0 Sodium Glycerophosphate 27.0 Magnesium Chloride 4.3 Dextrose 11.0 Mannitol 118.0 Hydroxyethyl Starch 50 g/l TABLE III. Total Water-Loss Associated With Glycerolization Of The Cat ____________________________________________________ Animal Pre-Perfusion Post-Perfusion Kg./ % Lost As # Weight Kg. Weight Water Dehydration FGP-1 4.1 3.6 2.46 18 FGP-2 3.9 3.1 2.34 34 FGP-3 4.5 3.9 2.70 22 FGP-4 6.0 5.0 3.60 28 FIGP-1 3.4 3.0 2.04 18 FIGP-2 3.4 3.2 2.04 9 FIGP-3 4.32 3.57 2.59 29 References In Preparation