X-Message-Number: 2907
Date: 17 Jul 94 23:38:26 EDT
From: Mike Darwin <>
Subject: SCI.CRYONICS  high glycerol concentrations

Regarding the use of high concentrations of cryoprotectant (in this case
glycerol) Bob states the following:


<<That cracking occurs more readily in the presence of concentrated glycerol
seems to imply--other things equal--that high concentrations are undesirable.
The push by Mike (and Alcor when Mike was there) toward very high
concentrations seems to be related (my impression) to the push toward
vitrification rather than freezing. But since vitrification is still
unperfected, it may be that  imperfect freezing is better than imperfect
vitrification. We'll see.>>

Alas, other things are NOT EQUAL.

The reason(s) why we went to progressively higher concentrations of glycerol

are, regretably, not nearly so theoretical (i.e., wishing to approach or achieve
vitrification) as Bob speculates.  The story goes something like this:


1) The choice of 3-4 M glycerol for cryoprotection was based on an extrapolation
from nature.  In the 1950s Audrey Smith observed that hamsters could tolerate
about 60% of their brain water being converted into ice and still recover
apparently neurologically intact (although neither she nor anyone else had done
studies to verify that learned behavior -- memory -- was intact in animals so
treated).  Conversion of 60% of the water in a mammal's brain to ice seemed to

be compatible with neurlogical recovery and this volume % of conversion of water
to ice in the brain became known as "Smith's Criterion."

It is important to realize that Audrey Smith cooled her hamsters to a core
temperature of about -0.5C and brain core temperature of maybe about -0.75C to

-1.0C -- at those temperatures tissue will have about 60% of its water converted
to ice at the one hour mark (Smith's hamsters did not typically survive more
than one hour of freezing, longer periods were lethal).

Of course, we are going to be cooling cryopreservation patients to a far lower
temperature, in fact cooling them to -196C.  However, for practical purposes in
terms of the amount of ice formed, the changes are all completed by far high
temperatures -- say around -50C to -60C.  So, the question was, looking at a
phase diagram for glycerol-water solutions, how much glycerol do you need to
satisfy Smith's criterion of converting no more than 60% of the brain's volume
into ice?  The answer, about 4M glycerol.  We started perfusing people with 4M
glycerol.  This was documented in a paper Leaf, Darwin (Federowicz) and Hixon
entitled Case report: Two consecutive suspensions, a comparative study in
experimental human suspended animation (Cryonics  (6(11), 13 (Nov 1985)).

2) Starting in the early 1980's  we decided to see just how well we were doing
using both ischemic and nonischemic animals in a "cryonic suspension" model:
employing glycerol perfusion to 3-4M and freezing to -196C and rewarming and
subsequent evaluation by light and electron microscopy.

What we found was not good.  There was massive injury to brain  -- so much so

that the experts I had look at it called it a "tissue homogenate."  While it was

not quite that bad, it was pretty bad.  The liver in these animals (cats) looked
even worse (it truly was a tissue homogenate!).  However, by contrast the heart
and  kidney were surprizingly well preserved -- still severely  damaged, but at
least the overall architecture of the tissue was intact and much of the damaged
structure seemed "inferable.". This seemed paradoxical. Meanwhile, Greg Fahy's

whole brain glycerolization and freezing experiments pointed to less injury with
high concentrations of agent.  He recommended that we go to 5M glycerol.

3)  Further studies were undertaken by Greg Fahy a few years later.
Greg used a perfusion model employing rabbits perfused to 3.67M glycerol but as
opposed to thawing the tissue, fixing it and evaluating it, he cut  it into
blocks at -140C (using a band saw), fractured out brain tissue away from the

cuts, and freeze-substituted and fixed it at -79C.  This technique allows one to
look at how and where the ice has actually formed to damage the tissue.

The results of these studies were shocking.  I hate to use a word like that
because it is so loaded and unscientific, but I am at a loss to find a more
descriptive one.  I believe I can justify the use of that word by using lots of
others to describe the results of these studies.

There was massive and large ice crystal formation in the brain.  A lesion which
was very frequent in my cat EMs (after thawing) was the presence of gaping
pericapillary holes and frequent 10-30 micron tears in the neuropil at 30-50
micron intervals.  These lesions were also seen in the frozen state: they were
caused by clearly visible masses of ice.  But of greater concern still was the
overall large-scale compression and distorition of the tissue in almost every
plane on both an ultrastructural and a histological level.  Furthermore, long

processes (axons, dendrites) were cut with great frequency in every direction on
the histological level (it was less possible to tell this at the EM level
because the field of view is so much smaller).

That's how we first arrived at 4M glycerol and how we went  to 5M.  We knew
nothing about cracking at that time.  True, we had observed it in solutions of
glycerol-water which could be vitrified (i.e., 72% glycerol) but it did not
occurr to us it was happening in biological systems subjected to freezing with
subvitrifiable concentrations of agent present.

On the basis of the freeze-substitution studies we decided to experiment with
higher concentrations of glycerol: up to 6M.  We found this concentration to be
perfusable, and we found it to markedly reduce the ice injury.  We knew about
fracturing by this time.

ereg then began to undertake a complex and reasonably exhaustive series of
experiments with rabbit brain slices evaluating a wide variety of
cryoprotectants and mixtures of cryoprotectants.  So far, as of this writing,

glycerol still appears to be best, although at my urging he has (a few days ago)
completed an experiment using vitrification solution.  High concentrations of
glycerol (approaching 7M) were far more protective histologically and
ultrasructurally) than lower concentrations.

Thus, we had a choice:  do we want to accpt massive ultrastructural and
histological injury and forgoe macroscopic cracking (even though we knew that
microscopic fracturing would *still* occur with lower concentrations of agent)
or did we want to further suppress the ice injury and accept (temporarily, at
least until intermediate temperature storage could be developed) gross cracking
injury?  We chose the latter.  And we did this only after careful consultation

with a wide variety of knowledgeable people such as Greg Fahy, Ralph Merkle, and
Eric Drexler.

Did we make the right choice?  Well, as Bob says, "we'll see."


Part of my problem with Bob's postings is that they simply do not acknowledge
that all this work has been done and that this informations EXISTS.  Reports of

this data have been published and presented at cryonics conferences and symposia
for nearly a decade.  In fact, my first papers on the ultrastructural and
histological preservation of cat brains were made at the Lake Tahoe Life
Extension Festivals in the mid-1980's.  Videotapes of these presentations were
available from the Festival (Fred and Linda Chamberlain) and the videos were
more than adequate to show the lesions displayed on the 35 mm slides being
projected and the extensive discussion of materials and methods which

accompanied them.  I know that Alcor still has copies of these tapes and I would
not be surprized if both Alcor and the Chamberlains would still be willing to

make them available.  If they are, I would very much like a copy -- I don't have
one myself!

The point is, no one from CI apparently viewed these tapes, attended  the

presentations, or even asked for  the details.  Ditto for my Cryonet posted copy
of the paper describing this work!  In fact, no one but Thomas Donaldson and
Mike Perry have ever even asked to *look* at the EMs of the brain and other
organs from this study.


This total lack of interest left me more than a little disheartened.  In fact, I
concluded that most people just don't care.  They are quite content to let the
future sort out all the problems.

Now, I hardly mean to represent my cat data as the be-all-end-all.  In fact, it
is a woefully inadequate study with many caveats.   I didn't realize how
inadequate till I started writing it up a number of years ago.  But it still
deserves to be looked at, and the fact is that some of the most troubling
lesions we observed were subsequently observed by another, independant
investigator.

And incidentally, I also have a collection of EM's from straight frozen brains
and brains frozen with low concentrations of glycerol.  If you think that's the
better approach, forget it!  A tissue homogenate *is* the operative description
in these cases (at least after thawing -- an important caveat).


One major difference in the CI technique from what is used in humans (I presume)
by CI and from is used in humans elsewhere is that as I understand it CI
*uniformly* cut a large window in the skull.  This may have eliminated or
reduced the "container effect" which we have all observed in the past.  For
instance it is possible to cool *isolated* brains loaded with high
concentrations of cryoprotectannt  to -196C without cracking if you peel them
away from the container they are in before cooling below the glass transition

point.  The work being done by the Russians should definitely look at some heads
where the brain case has not been opened in this fashion.

In humans that I am preparing for cryopreservation, if there is  cerebral
shrinkage, I am now making it a point to suction off all the fluid from the
cranial vault so that brain is mostly out of contact with its container in the
hopes that will reduce,  if not eliminate cracking.

 Also, we can hope against hope that very slow cooling rates *are* the
protective thing in the sheep heads and this will be applicable to humans.


Finally, Bob notes:

<<MIke says Terri Cannon was cooled from dry ice to liquid nitrogen temperature
in about 12 days, whereas we usually take only 7 days. However, he also says
she was cooled to dry ice temperature in only two days, whereas we take 7
days for this phase also.>>

This is true, and there is good reason for it.  For one thing partients loaded
with multimolar concentrations of glycerol will freeze at subzero temperatures.
It is also important to cool them to their freezing point and reduce their

temperature reasonably rapidly to inhibit biochemical/autolytic changes going on
at higer temperatures.  As I understand it, Bob places his patients after
perfusion in a sleeping bag and then gas cools them by slowly lowering them in
LN2.  My suspicion is that these patients are spending many, many hours (if not
days) above, at, or near their freezing temperatures.  The reason for this
assumption is that most of heat that has to be removed is the latent heat of
fusion when the water freezes.  An insulated patient in an air cooled
environment will present a big load of calories to dissipate leaving the
patient's core temperature hanging at the freezing point for a long time.  This
is why we (BPI and Alcor) use a liquid bath because it gets us over this "hump"
and moves the heat rapidly (i.e., 3-4C /hr) at relatively high subzero
temperatures. Temperatures at which biochemical activity is still very active.
We ran into this problem many years ago with the Berkowitz cryopreservation; he
was placed in a container thinly wrapped in fiberglass insulation and his
cooling rate was *very* slow.

Furthermore, while I could be wrong, I do not think there is any evidence that
the rate of cooling to temperatures considerably above Tg has any effect on
subsequent fracturing. Certainly both our experienmts with bulk solutions and
Greg Fahy's work do not bear this out.  Reasonably rapid cooling to at -60C
would seem prudent.

Whole body patients of course do take longer to reach this temperature -- about
3 days in a liquid bath.  I would be very interested in seeing specific data
from Bob's patients.  I have the following questions:

1) How are CI patients prepared for cryopreservation?  How is the perfusate
prepared/stored, how is perfusion carried out (temperatures, pressures, flows,
etc).  Ditto for transport procedures.

2) Can you post or send to interested parties copies of cooling curves for CI
patient(s) which show at least surface versus core temperatures?  

3) Are such data available for the sheep head work?

The availability of this specific data would be very helpful in comparing notes
and determining what is important.

Finally, what we have here as much as anything is a lack of communication.  The
band width has been very narrow or even nonexistent.  I am not concerned about

making an issue of why this has been so.  I'll be happy to accept any reasonable
share of the responsibility that is mine.  Much more to the point is *fixing*

the problem.  There is going to be a conference in November.  I think this would
a very good opprtunity for everyone to get together and discuss these issues.
And if the September conference is too soon or inappropriate, perhaps another
meeting, exclusively focused on ischemic and cryopreservation injury could be
set up -- perhaps somewhere more centrally located in the United States
(Detroit, Indianapolis, St Louis?).  I think we could all benefit by sharing
information and by talking in detail about where we go from here.  We each have
unique strengths in our capabilities.  It would be very nice to exploit those
strengths -- including those available in Russia, to our mutual advantage.
What about it?

Mike Darwin

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=2907