Message: #2631 - BPI Tech Brief #3
Date: 05 Mar 94 02:36:05 EST
From: Mike Darwin <>
Message-Subject: SCI.CRYONICS  BPI Tech Brief #3



        Steven B. Harris, M.D. and Michael G. Darwin

     This article is not about the last moments of Alexander Hamilton.  
A "burr-hole" is a standard medical term for a small window in the 
skull made by a surgeon.  Such a "craniotomy" or skull penetrating 
surgery is today routinely performed on cryonics patients in order to 
evaluate the adequacy of blood washout/cerebral perfusion, and 
ascertain the development of cerebral edema in a timely fashion during 
cryoprotective perfusion.  It has historically been very useful for 
this purpose, but has had a side effect which is our subject.  

    A typical burr-hole used to assess a cryonics patient is circular 
window in bone some 5-10 mm (1/5 to 2/5 inches) in diameter, and it is 
opened near the midline over the parietal or frontal cortex of the 
brain soon after the patient has been brought to the operating table 
for cryoprotective perfusion. This happens some 3 to 12 hours, 
typically, after legal death has been pronounced, depending on 
transportation time from the area of the U.S. in which the clinical 
death of the patient occurred.  The usual procedure for a burr hole is 
toincise the scalp in the top of the head with a #10 scalpel blade, 
making an incision 3-5 cm in length.  The "periosteum" tissue covering 
the bone of the skill is then incised with the scalpel and reflected 
with a periosteal elevator.  Craniotomy or penetration of the skull is 
then carried out using either a pneumatic perforator or a Hudson brace 
with Cushing burr (a manual drill much like a standard hand wood 

     This procedure, surprisingly, is relatively simple to do without 
injury to the brain.  Next, the "dura mater," which is the thickest 
and toughest ("durable") of the layers of tissue protecting the brain 
and spinal cord, and the one which is outermost, is opened.  For this, 
one usually uses a dura hook to retract the dura away from the brain 
so that it is not cut when incised with sharp tip of a #11 scalpel 
blade. The dura flaps are then trimmed away to the margins of the 
opening in the bone.

     This procedure was first implemented in January of 1980 on a 
patient prepared for Trans Time, Inc. by Cryovita Laboratories 
(Cryonics (6(11), 13 (Nov., 1985).  The procedure allowed for good 
visualization of the pial vessels (vessels in the pia mater, the thin 
membrane next to the brain tissue itself).  It also allowed good 
observation of the degree of blood washout of these vessels in a 
patient that had been subjected to air transport packed in ice in the 
absence of total body washout (i.e. with blood still present in the 
body). Subsequently this technique has been used on all patients 
subjected to cryoprotective perfusion by the Alcor Foundation and has 
been employed also more recently by Biopreservation.  It has proven 
very useful in indirectly evaluating the degree of cerebral capillary 

    In patients transported under optimum conditions, with minimal 
near death "agonal" ischemic insult due to shock (in this context 
tissue underperfusion), the patient's brain typically becomes 
dehydrated in response to glycerolization, and remains dehydrated 
throughout the course of cryoprotective perfusion, with brain volume 
losses of 30 to 50% of baseline volume being common by the termination 
of perfusion as estimated by measuring retraction of the cerebral 
hemispheres.  By contrast, cryonics patients with severe pre-perfusion 
ischemic injury (typically caused by a delay between clinical death 
and when access to the patient was permitted to cryonicists) either do 
not exhibit cerebral dehydration and shrinkage, or else they exhibit 
it transiently, followed by the development of progressive and massive 
cerebral edema. Perfusion is usually terminated when the brain begins 
to herniate (bulge) significantly into the burr opening.  In 
cryopreservations performed where pronouncement of death is timely and 
access to the arrested patient almost immediate, edema typically does 
not force termination of perfusion.  In perfusion of patients 
suffering long periods of ischemia before access was permitted, brain 
edema as monitored by the burr-hole has typically forced termination 
of perfusion.

   A puzzling phenomenon which has been observed fairly consistently 
in patients who have exhibited marked cerebral dehydration during 
glycerolization (i.e., those thought to have the least brain insult 
and the best suspensions), is the leakage during cryoprotective 
perfusion of comparatively large flows of perfusate from the 
craniotomy opening, usually in the amount of 150 to 250 cc/min.  It 
has been assumed until recently that all of this leakage was from the 
incised bone, scalp and dura.  Since these tissues are normally cut 
following blood washout, normal intraoperative hemostasis would not be 
expected to occur (with no blood there is no clotting).

      Over the course of the years one of us (M.G.D.) has made a 
number of efforts to control this leakage as well as determine its 
source.  This was something of a priority, since the loss of this 
volume of perfusate from the recirculating system, particularly in the 
case of neuropreservation patients, constitutes a volume roughly equal 
to that typically withdrawn (and added) to achieve the linear increase 
in glycerol concentration in the recirculating perfusate.  In 
neuropreservations it also constitutes a significant fraction of the 
total recirculating and concentrate perfusate volumes over the course 
of a typical 2-3 hour cryoprotective perfusion.  Additionally, unless 
this perfusate is recovered and returned via cardiotomy suction to the 
recirculating system, it constitutes not only a hard to quantify 
"unknown" affecting the rate of glycerol concentration increase, but 
also a significant housekeeping problem as the perfusate is lost to 
the table top where it can saturate drapes  and complicate operating 
room and patient clean-up following the perfusion (this is of 
particular concern in patients with hepatitis, HIV or other infectious 
blood borne disease).

      Early efforts at fluid loss control consisted of using bone wax 
to secure hemostasis in bone cortex,  and the use of clips and cautery 
to stop leakage from the scalp and dura.  None of these efforts was 
successful, and so it became increasingly apparent that the source of 
the leakage was intracranial, excluding the dura.   What was not clear 
was where this flow was coming from.  Did this leakage represent 
transudation (direct pressure leakage) from the pia-- the delicate 
membrane covering the brain?  Or did it come from perivascular spaces, 
having leaked from vessels?  Was damage to the choroid plexus, the 
normal source of cerebrospinal fluid, the source? 

    None of these explanations seemed probable.  Transudation and 
increased capillary permeability seemed more likely events in patients 
with greater rather than less ischemic injury, yet little that was 
consistent with this was noted.  In fact, drainage from the burr hole 
was routinely highest in patients transported under good conditions 
and with maximum amounts of cerebral dehydration!  By contrast, 
patients with poor cerebral perfusion who did not develop cerebral 
dehydration, or who developed only modest dehydration followed by a 
rebound to cerebral edema, exhibited little or no burr hole drainage.  
None of these facts fit very well.  Also, there seemed little reason 
why transudation, primarily expected to be a subarachnoid problem, 
would show up in the subdural space even before the arachnoid was 

      During the recent cryopreservation of American Cryonics Society 
patient ACS 9577, these troublesome problems were illustrated well.  
Once more, an attempt was made by M.G.D. to determine the source of 
the craniotomy drainage.  The scalp and periosteum were incised as 
usual and the bone was perforated using a DePuy pneumatic perforator.  
However, the dura was not opened at the start of cryoprotective 
perfusion.  Rather, cryoprotective perfusion was commenced and 
perfusate leakage from the scalp and bone were evaluated and 
determined to be no more than 10-15 cc/min.  Before the start of 
perfusion the dura was gently depressed with a probe and was found to 
be flaccid with the cerebral cortex not palpable at a depression of 1-
2mm.  The pressure of fluid in the subdural space was obviously not 

      Approximately 30 minutes after the start of perfusion the dura 
was again depressed and was found to be moderately tense.  Fluid 
apparently had begun to accumulate in the subdural space between the 
outer membrane protecting the brain (the dura), and the arachnoid 
membrane beneath, even though the outer membrane was at that point 
intact and there were no cuts as yet beneath to leak from.   Seventy-
three minutes into cryoprotective perfusion the dura was pierced with 
a 16 gauge needle and a copious, moderately high pressure flow of 
clear fluid was observed to issue from the needle.  At this point the 
needle was withdrawn and the puncture in the dura was widened with a 
#11 scalpel blade approximately 1 mm in diameter.  A copious and 
pressured flow of fluid was observed streaming from the puncture.  

     The flow rate of fluid out of the puncture in the dura was 
measured by collection in a graduate over a 1-minute period and was 
found to be 150 cc/min.  A nearly maximal fluid leak was occurring in 
deeper layers of the brain, obviously independently of surgical 

     The opening in the dura was then widened and a probe was passed 
to determine the degree of cerebral dehydration .  The 
cortical/arachnoid surface was determined to be 7 mm below the inner 
surface of the cranial vault.  A length of plastic tubing was then 
passed into the craniotomy such that flow from the dura, bone and 
scalp were excluded.  Flow  of fluid from the cranial vault was 
measured at approximately 140 cc/min.  This fluid was then evaluated 
for cryoprotective agent concentration and for blood gases and 
electrolytes.  The character of this drainage was venous in nature.  
This fact agrees with previously measured concentration of glycerol in 
burr hole drainage in both published (Cryonics 1986 Feb;7(2):15-32) 
and unpublished Alcor cryopreservation case histories A-1133 and A-
1169 in which burr hole glycerol concentration as measured by 
refractive index was found to overlap the measured glycerol 
concentration in the patient's venous (as opposed to arterial) 

Possible Sources of Intracranial Fluid

      The determination of the character of this drainage as primarily 
of venous origin deep to the dura and independent of surgical trauma, 
taken with consideration of the anatomy of the meninges, suggests 
three possible sources, any of which may operate alone, or any 
combination of which may operate together. 

     The most likely possibility is fluid leakage from tears or 
ruptures in the bridging veins between the dura and the next innermost 
membrane (the arachnoid), as a result of shrinkage of the cerebral 
hemispheres in response to glycerolization.  A second possibility is 
that shrinkage of the brain may partly separate the two layers of the 
dura mater (the endosteal which lines the inside of the skull and the 
meningeal layer which is deep to this, thereby disrupting one or more 
the large venous sinsuses which run between these layers in many parts 
of the skull, and which are not even properly vessels.  It is even 
possible that some past burr holes, perhaps placed too near the 
midline, have penetrated the endosteal layer of dura only to open into 
the large sagital venous sinus which runs down the midline at the top 
of the skull.  When not filled with blood such a sinus would appear 
only as one more potential space in the skull.   A third and final 
possibility is that loss of competency in the microscopic valves of 
the arachnoidal villi has occurred as a result of glycerol-induced 
dehydration of the endothelial cells which cover the villi, and which 
normally regulate the flow of cerebrospinal fluid into the venous 

      Localized injury to bridging veins from deceleration injury is a 
known source of intracranial venous bleeding and is a common cause of 
subdural hematoma, which results often from venous leakage into the 
subdural space.   This phenomenon has also been well described in 
nontraumatized patients receiving mannitol, an osmotic and dehydrating 
agent, during heart bypass (Surg Neurol 1985 Nov;24(5):520-524).   
Cryonics perfusate contains both mannitol and glycerol in large 
quantities.  The pronounced cerebral dehydration which occurs as a 
consequence of the perfusion of osmotically active agents during 
cryoprotectant perfusion necessarily results in separation of the 
arachnoid membrane from the dura (the dura remains adherent to the 
inside of the cranial vault) and presumably could also result in 
tearing of the bridging veins which connect the dura and the arachnoid 
membrane.  Hypothermia may contribute to reduced elasticity in such 
veins, and lack of hemostasis and clotting insures that if there is a 
rupture, it will leak fluid copiously and continuously.

      A second source of the fluid leak might be the superior sagital 
venous sinus, or one of several other venous sinuses which might 
potentially be disrupted by brain dehydration and separation of the 
two layers of dura which form them.  More care will be required in the 
future to place burr-holes off midline, and to attempt to identify 
both layers of dura when dura is being penetrated.

    A third fluid source might (though implausibly) be the arachnoidal 
villi.  The arachnoidal villi are microscopic projections of the 
arachnoidal membrane through the walls of the venous sinuses (large 
venous reservoirs which collect blood from brain flow).  Electron 
microscopy of the endothelial cells covering these projections 
discloses the presence of large vesicular holes which pass through the 
body of the cells.  These holes or pores are large enough to allow for 
the relatively free passage of the cerebrospinal fluid (CSF) into the 
venous blood, and may be large enough to allow for even the passage of 
some formed elements such red blood cells (Guyton AC, Textbook of 
Medical Physiology, W.B. Saunders, Philadelphia, 1991: 682-683.).  
Under normal physiologic conditions the arachnoidal villi act to 
reabsorb CSF when the pressure of the CSF is about 1.5 mm Hg greater 
than the pressure in the venous sinuses.  Normally the amount of CSF 
absorbed during the course of a day by this mechanism is modest, in 
the range of 150 to 200 cc.  The effect of osmotically active 
compounds such as glycerol, dimethylsulfoxide or other  cryoprotective 
compounds on the cellular "valving" mechanism in the arachnoid villi 
is unknown.  However, the possibility exists that cellular dehydration 
may greatly increase the size of the pores in the villi cells 
resulting in a leakage of venous effluent retrograde from its normal 
path, from the venous sinuses into the subarachnoid space.  

    Ordinarily the subarachnoid space in the brain does not 
communicate with the subdural space above it,  so that in the absence 
of an arachnoid membrane tear, a fluid leak in the subarachnoid space 
seems less likely to show up as a spontaneous and rapid subdural (not 
subarachnoid) fluid collection, which examination of the most recent 
suspension strongly suggested was the primary problem.  

Significance of The Problem

     While this kind of injury would be of great concern under 
physiologic conditions it is of significance to the cryopreservation 
patient only if a burr-hole is *not* opened in the skull.  In such a 
situation the accumulation of venous perfusate at venous pressure 
might significantly reduced brain perfusion, in effect creating a 
large bilateral subdural hematoma.

Determining The Source of The Fluid

     In subsequent cases where ischemic time is minimal and brain 
perfusion is good (with associated cerebral dehydration) we will 
undertake to definitively determine the source of the leakage. If we 
can identify the subarachnoid space during cryoprotective perfusion we 
will attempt to pass a needle into it to obtain some (subarachnoid) 
CSF.  If this is still chemically very different from venous return 
(and it should be in every scenario except the leaky villi one), then 
we can be fairly confident that the source of the leakage is subdural 
either from injury to bridging veins or torn venous sinuses.



AU  - Giamundo A ; Benvenuti D ; Lavano A ; D'Andrea F
TI  - Chronic subdural haematoma after spinal anaesthesia. Case report.
AB  - In this study an interesting and not frequent case of
      non-traumatic chronic subdural haematoma after spinal anaesthesia
      is reported. After a careful review of the cases described in the
      literature, the Authors discussed the physiopathological
      mechanisms, emphasizing how the break of the bridge veins or of
      the subarachnoid granulations, the cerebral atrophy and the
      dehydration are factors which facilitate the appearance of this
      neurological complication. They suggest that a suitable
      post-operative rehydration is an important prevention factor.

SO  - J Neurosurg Sci 1985 Apr-Jun;29(2):153-5
AU  - Yokote H ; Itakura T ; Funahashi K ; Kamei I ; Hayashi S ; Komai N
TI  - Chronic subdural hematoma after open heart surgery.
AB  - Three cases of chronic subdural hematoma after open heart surgery
      are reported. In all cases, computed tomography scans revealed
      subdural accumulations of high density after cardiac surgery. The
      high-density areas changed into isodensity or low density with
      mass effect within 2 or 3 weeks. Anticoagulant (heparin) and a
      tearing of bridging veins after a rapid change of the brain
      volume by administration of mannitol can be a cause of chronic
      subdural hematoma. Forty-five to 60 mL of liquefied hematoma was
      aspirated and the outer membrane of the hematoma cavity was
      recognized by a trepanation.
SO  - Surg Neurol 1985 Nov;24(5):520-4