Epidu al egional hypo he mia o p e en ion
o pa aplegia a e ao ic occlusion:
Expe imen al e alua ion in a abbi model
Jose A. Gonzalez-Faja do, MD, Bea iz Agui e, MD, Jose L. Pe ez-Bu kha d , MD,
Tomas Al a ez, MD, I~opoldo Fe nandez, MD, GuiUe mo Ramos, MD, and
Ca los Vaque o, MD,
Valladolid, Spain
Pu pose:
The e icacy o epidu al egional hypo he mia in he p e en ion o acu e and
delayed-onse pa aplegia, as well as possible complica ions and limi a ions o his
echnique o a clinically accep able o m, we e e alua ed in 49 New Zealand whi e abbi s.
Me hods:
A modi ied abbi spinal co d ischemia model o in a enal ao ic occlusion o
30 minu es was employed. The s udy was pe o med in wo phases. In phase I (n = 20),
egional hypo he mia induced by epidu al pe usion o iced no mal saline solu ion (4 ~ C)
was es ed e sus con ol in 10 abbi s each (g oups A and B). In phase II (n = 29) he
animals we e subdi ided in o h ee g oups o s udy he kine ics o abso p ion and
dis ibu ion o me hylene blue (g oup C; n = 10), adiog aphic con as ma e ial (g oup
D; n = 9), and measu emen o ce eb ospinal p essu e while an epidu al iced solu ion was
o was no in used (g oup E; n = 10).
Resul s:
A 24 and 48 hou s, all o he no mo he mic animals showed i e e sible
pa aplegia (Ta lo sco e 0). In con as , a 24 hou s none o he abbi s unde going
epidu al cold in usion we e pa aplegic, al hough a 48 hou s one animal had weakness o
a hindiimb (Ta lo sco e 3). Plasma concen a ion- ime p o iles o a con inuous epidu al
pe usion wi h me hylene blue showed ha he spinal canal is a highly complian space.
Epidu og aphs showed ha epidu al pe usion ends o sp ead mo e in a cephalic han
caudal di ec ion and he main up ake is by he ascula compa iiien . Despi e he la ge
olumes in used (78.75 ml/h ; ange, 50 o 100 ml), we obse ed only a modes ansien
inc ease in ce eb ospinal luid p essu e ( om 2.5 -+ 0.3 mm Hg o 5.4 -+ 0.1 mm Hg),
al hough some animals had in ac anial hype ension.
Conclusions:
Regional hypo he mia induced by epidu al cold pe usion has a highly
p o ec i e e ec agains he ischemic spinal co d damage. Howe e , his me hod p obably
does no a oid he isk o delayed-onse pa aplegia. An impo an limi a ion o his
echnique is he di icul y o con olling he in a hecal p essu es. (J VAsc SURG
1996;23:446-52.)
The mos de as a ing and unp edic able compli-
ca ion a e su ge y o he descending and ho acoab-
dominal ao a is pa aplegia. This complica ion anges
om 0.5% o 38%, ~ depending on ac o s such as he
ype and ex en o econs uc ion, he p esence o
F om he Di ision o Vascula Su ge y, Hospi al Uni e si a io,
Valladolid.
This wo k was awa ded he E. Be en P ize a he 5 h Con en ion
o Spanish-speaking Vascula Su geons, San Juan, Pue o Rico,
Oc . 11-13, 1995.
Rep in eques s: Jose A. Gonzalez-Faja do, MD, Di ision o
Vascula Su ge y, Hospi al Uni e si a io, 47011 Valladolid,
Spain.
Copy igh 9 1996 by The Socie y o Vascula Su ge y and
In e na ional Socie y o Ca dio ascula Su ge y, No h Ame i-
can Chap e .
0741-5214/96/$5.00 + 0 24/1/68148
446
dissec ion, and he du a ion o ao ic c oss-clamping.
Se e al adjunc i e echniques and d ugs ha e been
used o p o ec ing he spinal co d, bu none o hese
measu es has comple ely p e en ed pa aplegia. 2-1z
Induced hypo he mia, ei he sys emic o local,
seems o p o ide he mos po en p o ec ion agains
ischemia- epe usion neu al inju y) ,13w Hypo he -
mia has been shown o educe he oxygen consump-
ion o neu al issue by app oxima ely 6% o 7% o
each deg ee o co e empe a u e educed and esul s
in an inc ease in sa e ischemia ime o he spinal co d
by app oxima ely 5 o 6 minu es o each deg ee
Celsius o empe a u e educ ion) s Sys emic hypo-
he mia ca ies a isk o ca diac diso de s and
coagulopa hy.19,2o Regional spinal co d hypo he mia
has been ob ained expe imen ally by pe using he
JOURNAL OF VASCULAR SURGERY
Volume 23, Numbe 3
Gonzalez-Faja do e al. 447
essels supplying he co d wi h cold blood o
c ys alloid 2ms o by pe using he suba achnoid 24'2s
o epidu a126-3~ space wi h hypo he mic solu ions.
The esul s show ha hypo he mia p oduces a
a o able ou come wi h ega d o neu ologic de ici s;
howe e , some imes cooling is no con olled and
sys emic hypo he mia may de elop. O he imes i s
clinical applica ion is limi ed. Recen ly Da ison e
al.3~ epo ed hei clinical expe ience in eigh pa-
ien s wi h a echnique o selec i e spinal co d
hypo he mia by epidu al cooling. They showed ha
he epidu al cold pe usion was a sa is ac o y me hod
o achie ing egional spinal co d hypo he mia
( ange, 25 ~ o 28.8 ~ C).
Gi en ha he spinal co d cooling can p o ide
p o ec ion om ischemic spinal co d inju y du ing
he c i ical pe iod o ao ic c oss-clamping and
in e cos al essel eanas omosis, he pu poses o his
s udy ha e been, i s , o e alua e he e icacy o
epidu al egional hypo he mia in a abbi model o
ao ic occlusion, wi h special a en ion o acu e and
delayed-onse pa aplegia and, second, o de e mine
possible complica ions and limi a ions o his ech-
nique o a clinically accep able o m.
MATERIAL AND METHODS
Fo y-nine New Zealand whi e abbi s o bo h
sexes (2.0 o 3.5 kg) we e used o his s udy. A
modi ied abbi spinal co d ischemia model o
in a enal ao ic occlusion o 30 minu es was em-
ployed. All expe imen s we e conduc ed in compli-
ance wi h he "P inciples o Labo a o y Animals
Ca e" o mula ed by he Eu opean Council and
published by he Na ional Ins i u es o Heal h
(Spain, B.O.E. No. 67: 8509-12, 1988).
S udy phases
The s udy was pe o med in wo phases. In phase
I (n = 20), epidu al egional hypo he mia was es ed
e sus con ol in 10 abbi s each (g oups A and B).
A o al o 30 minu es o in a enal ao ic occlusion
was pe o med ollowed by s ic neu ologic exami-
na ion a 24 and 48 hou s. Neu ologic ou come was
e alua ed acco ding o Ta lo 's c i e ia32: g ade 0,
spas ic pa aplegia and no mo emen o he lowe
limbs; g ade 1, spas ic pa aplegia wi h sligh mo e-
men o he lowe limbs; g ade 2, good mo emen o
he lowe limbs bu unable o s and; g ade 3, able o
s and bu no able o walk no mally; and g ade 4,
comple e eco e y.
In phase II (n = 29) he abbi s we e subdi ided
in o h ee g oups: (1) animals ma unde wen a
0.25% solu ion o me hylene blue by epidu al
egional pe usion (g oup C; n = 10); (2) animals
ha unde wen adiog aphic con as s udy by epi-
du al pe usion (g oup D; n = 9); and (3) animals
wi h measu emen o ce eb ospinal luid (CSF)
p essu e while an epidu al iced no mal saline solu ion
was o was no in used (g oup E; n = 10). The
kine ics o abso p ion and dis ibu ion o me hylene
blue and adiog aphic con as ma e ial we e de e -
mined by spec oscopic me hods 33 and adiog aphy,
espec i ely. The ela ionship be ween CSF p essu e
and epidu al pe usion was analyzed.
Su gical p ocedu e
The animals we e anes he ized wi h in amuscu-
la ke amine hyd ochlo ide (40 mg/kg). In aope a-
i e ac ional doses we e used o main ain app op i-
a e le els o anes hesia and p e en he need o
endo acheal in uba ion and mechanical en ila ion.
The animals we e immobilized on he ope a ing
able, and he u in he skull, neck, lumba spine, and
abdomen was clipped wi h elec ic shea s, and he
skin was p epa ed wi h iodine solu ion. A e local
in il a ion o mepi acaine hyd ochlo ide (10
mg/kg), a midline incision was pe o med a L4-L5.
The muscula u e and he spinal p ocess o he ou h
lumba e eb a we e emo ed. A small incision was
made in he in e e eb al ligamen . A cannula
(Epidu al Minipack, 18 gauge; Po ex L d.; Hy he,
Ken , U.K.) was inse ed in o he epidu al space
abou 5 cm c anially and hen ixed and he wound
was closed wi h eabso bible su u e.
In he supine posi ion a la e oce ical incision o
he neck was made o allow iden i ica ion o he
ce ical essels: Two 20-gauge hepa inized saline-
illed ca he e s we e inse ed in o he jugula ein and
ca o id a e y. An in a enous in usion o iso onic
saline solu ion a a a e o 1 ml/min was adminis e ed
du ing he ope a ion o eplace su gical losses.
Con inuous on-line moni o ing o sys emic blood
p essu e was main ained h ough he ca o id ca he e
(S a ham p essu e ansduc o ; Viggo-Spec amed,
Inc., C i ical Ca e Di ision, Oxna d, Cali .). A e
local in il a ion o mepi acaine hyd ochlo ide (10
mg/kg) a he abdominal alba line, he in a enal
ao a was app oached by a xiphopubic midline
lapa o omy. A e sys emic hepa iniza ion (1.5
mg/kg in a enously), he ao a was clamped o 30
minu es. Fo he body empe a u e measu emen , a
lexible empe a u e p obe was placed in he ec um.
The di ec spinal co d empe a u e was no measu ed
in he hypo he mic animals. The e ec i eness o
epidu al cooling was assumed based on o he s udies
in he li e a u e. 27,sl
JOURNAL OF VASCULAR SURGERY
448 Gonzalez-Faja do e al.
Ma ch 1996
Epidu al egional
hypo he mia
An in usion o iced no mal saline solu ion (4 ~ C)
was begun in o he epidu al ca he e wi h he goal o
dec ease he empe a u e o he CSF. The in usion
was begun 30 minu es be o e clamping, main ained
du ing ao ic c oss-clamping (30 minu es), and
s opped a he end o he pe iod o ischemia.
Pe usion cooling s a ed a a a e o 1.4 ml/min and
was modi ied (inc eased o dec eased) depending on
he neu ologic ole ance o he animal. When he
in a enal ao a was undamped, he epidu al ca he e
was emo ed and he wounds su u ed. The animals
we e placed in cages o neu ologic e alua ion a 24
and 48 hou s.
S udy
o he abso p ion
and dis ibu ion
o he
epidu al
pe usion
Plasma me hylene blue concen a ions we e de-
e mined in 10 abbi s (g oup C) a e epidu al
adminis a ion o a 0.25% solu ion o me hylene blue
(1.4 ml/min). Venous blood samples we e collec ed
15, 30, 45, 60, and 70 minu es om he in e io ena
ca a (0.5 ml) by punc u e wi h an insulin sy inge and
he jugula ca he e (0.5 ml). Plasma was sepa a ed
and kep ozen un il analyzed by spec opho ome y.
To minimize e o s, all de e mina ions we e calcu-
la ed wice in each pe iod. The alue was exp essed as
he mean o each sample s udied.
In nine animals (g oup D) adiog aphic con as
ma e ial (Omnig a -300, Jus e, S.A.Q.F.; Sche ing
AG, Be lin, Ge many) was in used in o he epidu al
ca he e . The abbi s we e sepa a ed in o h ee
subg oups o h ee animals acco ding o he a e o
pe usion. Th ee animals ecei ed a bolus o 10 ml,
h ee ecei ed a 10 ml solu ion o 30 seconds, and,
inally, ano he h ee had an in usion o 10 ml o 1
minu e. Radiog aphy was pe o med a he begin-
ning, middle, and end o each pe usion.
Measu emen o CSF p essu e
CSF p essu e was measu ed in 10 abbi s (g oup
E) unde wo di e en condi ions: wi h and wi hou
he epidu al cold pe usion (con ol, n = 5; hypo-
he mia, n = 5). To measu e CSF p essu e (S a ham
p essu e ansduc o ; Viggo-Spec amed, Inc., C i i-
cal Ca e Di ision), he animals we e u ned wi h hei
shoulde up igh , exposing he pos e io aspec o he
neck and base o he skull. A e local in il a ion o
mepi acaine hyd ochlo ide (10 mg/kg), a pos e io
midline incision was made and an 18-gauge angio-
ca he e was placed in o he cis e na magna. The
cannula was ixed wi h den al cemen o a oid luid
leak. Co ec placemen was checked by obse ing he
wa e o m o he CSF p essu e, in pa icula looking
o espi a o y and ca dio ascula a ia ion.
His ologic
s udy
The animals we e killed a he comple ion o he
p ojec wi h a le hal in a enous dose o phenoba -
bi al and po assium chlo ide. The spinal co d (p oxi-
mal and dis al o he le el o ao ic occlusion) was
emo ed o ixa ion in 10% neu al bu e ed o ma-
lin o a leas 10 days. A e embedding in pa a in
wax, specimens we e sec ioned a 5 wm and he
sec ions we e s ained wi h hema oxylin and eosin.
S a is ical
analysis
The alues ob ained we e en e ed as a da a
ma ix in he s a is ical p og am SIGMA (Ho us
Ha dwa e, Mad id, Spain). Resul s a e shown as he
mean _+ SEM. Hemodynamic and empe a u e
compa isons we e analyzed by he pai ed o un-
pai ed S uden es . Neu ologic ou come in he
di e en g oups (Ta lo 's sco e) was compa ed wi h
Fishe 's exac es . S a is ical signi icance was as-
sumed a p < 0.05.
RESULTS
P oximal ca o id a e ial p essu e was simila in
all g oups a baseline (95 _+ 10 mm Hg in con ol
animals; 90 -+ 15 mm Hg in animals wi h hypo he -
mia). Wi h ao ic occlusion, a signi ican inc ease
(p < 0.05) was obse ed in p oximal p essu es in all
g oups o abbi s compa ed wi h espec i e baselines
(125 _+ 12 mm Hg and 115 _+ 11 mm Hg, espec-
i ely). No di e ences we e obse ed be ween he
no mo he mic and hypo he mic g oups. A e e-
mo al o he ao ic clamp, he p oximal blood
p essu e s abilized o a poin no signi ican ly di e -
en om he baseline alue.
In he no mo he mic animals, body empe a u e
did no change du ing ischemia. The ec al empe a-
u e o hypo he mic animals dec eased om he
con ol le el o 37.5 ~ _+ 0.6~ o 35.1 ~ _+ 0.8~ a
he end o he ischemic pe iod (p < 0.001).
Neu ologic
ou come
E alua ion o he neu ologic ou come is shown in
Table I. A 24 hou s all o he no mo he mic animals
(n--10) showed i e e sible pa aplegia (Ta lo
sco e 0). No di e ences we e ound a 48 hou s. In
con as , a 24 hou s none o he abbi s unde going
epidu al cold in usion we e pa aplegic, al hough a
48 hou s one animal had a weakness o he le
hindlimb (Ta lo sco e 3). The di e ence in Ta lo
sco e be ween animals unde going no mo he mic
spinal co d ischemia and hypo he mic pe usion was
signi ican (Fishe s exac es , p < 0.001). In addi-
ion, ou animals showed clinical signs o in ac anial
hype ension (exoph haimos) du ing he hypo he -
mic epidu al pe usion (phase I). In hese cases, he
JOURNAL OF VASCULAR SURGERY
Volume 23, Numbe 3
Gonzalez-Faja do e al. 449
Table I. Neu ologic ou come a 24 and 48 hou s
24 Hou s 48 Hou s
G oup A G oup B* G oup A G oup B ~
No. o abbi s 10 10 10 10
Ta lo '
sco e
0 10 0 10 0
1 0 0 0 0
2 0 0 0 0
3 0 0 0 1
4 0 10 0 9
G oup A, No mo he mic animals;
G oup B,
hypo he mic animals.
up < 0.001, Fishe 's exac es .
a e o in usion was s opped o dec eased o a oid he
dea h o he animals.
CSF p essu es
Th ee
animals wi h inco ec placemen o he
ca he e in o he cis e na magna and one wi h a
bloody spinal ap we e excluded om he s udy. The
CSF p essu e was simila in bo h g oups a baseline
de e mina ion (2.9 + 0.5 mm Hg and 2.5 + 0.3
mm Hg, espec i ely). Immedia ely a e ao ic
clamping, he CSF p essu e ose in bo h g oups
(4.77 _+ 0.4 mm Hg and 4.5 + 0.35 mm Hg). Wi h
he epidu al iced in usion, he CSF p essu e emained
in ac in h ee animals (5.4 _+ 0.1 mm Hg), bu a
abbi showed signs o in ac anial hype ension
(exoph halmos) and CSF p essu e o 13.23 mm Hg
a he end o he expe imen . S a is ical analysis o
CSF p essu e was no pe o med because o small
sample size (con ol g oup, n--2; hypo he mic
g oup, n = 4).
Abso p ion and dis ibu ion o he
epidu al pe usion
The mean olume adminis e ed was 78.75 -+ 5.0
ml/h ( ange, 50 o 100 ml/h ). A e 1 minu e o
epidu al pe usion, me hylene blue appea ed in he
e ope i oneal space in o he lympha ic duc s. De-
e mina ion o he plasma concen a ion- ime p o-
iles showed ha abso p ion a e epidu al pe usion
is biphasic: a apid ini ial abso p ion was ollowed by
a much slowe second phase (Fig. 1). The maximal
concen a ion (120.12 _+ 7.37 ~g/100 ml) was
eached a 60 minu es. The concen a ion- ime
cu e inc eased ollowing a linea ela ionship
(Y = 20.975 + 1.6186X; = 0.72341; p <
0.001). The spinal co d nec opsy a he end o he
expe imen showed ha he me hylene blue injec ed
in o he lumba epidu al space end o sp ead mo e in
a cephalic han caudal di ec ion. Likewise, he spinal
oo s and co d appea ed dyed, al hough de ailed
dis ibu ion wi hin CSF was impossible o de e -
mine. Th ee animals showed exoph halmos and
conjun i al edema wi h blue eyes.
On he o he hand, he epidu og aphs a e
injec ion o a adiog aphic con as agen co obo-
a ed he os al mo emen om he lumba space.
Vascula up ake occu ed apidly in o he epidu al
eins (Ba son's plexus). T anspo o gene al ci cu-
la ion was p edominan ly by he azygos collec ion
sys em and in e io ena ca a. The adiog aphs aken
a e 1 minu e o epidu al pe usion showed he hea
and bo h kidneys wi h con as ma e ial (Fig. 2). Fi e
animals died du ing he expe imen , wi h signs o
in ac anial hype ension: h ee abbi s unde going a
bolus o 10 ml and wo abbi s unde going 10 ml o
epidu al pe usion o 30 seconds. The o he ou
animals showed a di e se deg ee o in ac anial
hype ension (exoph halmos) ha disappea ed ap-
idly a e se e al minu es.
His opa hology
The his opa hologic indings clea ly co ela ed
wi h he neu ologic s a us. No mo he mic animals
showed o al des uc ion o an e io ho n cells wi h
p onounced acuoliza ion and gliosis. The spinal
co d s uc u e was well p ese ed in hypo he mic
animals ha we e spa ed neu ologic de ici . No
in a c ions we e de ec ed in his g oup. The animal
wi h delayed-onse pa esis o a hindlimb had an
in e media e deg ee o ganglion cell des uc ion wi h
hype ch oma ic nuclei and mild o mode a e acu-
oliza ion o cy oplasm. The animals om phase II,
killed immedia ely a e he expe imen , showed
some conges ed a eas and essels wi h abundan ed
blood cells.
DISCUSSION
The abbi model o ao ic occlusion was de-
sc ibed by Zi in and de Gi olami s4 and has been
es ed by o he au ho s ls'2s as a eliable model o
p oducing neu ologic de ici s om ischemic spinal
co d inju y, including acu e and delayed pa aplegia.
JOURNAL OF VASCULAR SURGERY
450
Gonzalez-Faja do e al.
Ma ch 1996
Fig. 1. Plasma me hylene blue concen a ion- ime p o ile
shows ha abso p ion a e epidu al pe usion is biphasic:
apid ini ial abso p ion is ollowed by much slowe second
phase. Concen a ion- ime cu e ollows linea ela ionship
(Y = 20.975 + 1.6186X; = 0.72341;p < 0.001).
Moo e and Hollie s5 also showed ha p olonged
spinal co d ischemia (mo e han 27 minu es) p o-
duced pe manen pa aplegia. Wi h his model, we
ha e e alua ed he in luence o an epidu al cold
pe usion on ischemic spinal co d inju y. This s udy
demons a es ha he egional hypo he mia induced
by epidu al cold pe usion has a highly p o ec i e
e ec agains ischemic spinal co d damage. In ac ,
epidu al cooling was associa ed wi h pos ope a i e
p ese a ion o neu ologic unc ion a e 30 minu es
o in a enal ao ic c oss-clamping, al hough one
animal had delayed-onse pa esis o a hindlimb a 48
hou s. These esul s sugges ha he mode a e
cooling eached by his me hod may be enough o
p o ide ull p o ec ion du ing he acu e ischemic
inju y bu no du ing he epe usion pe iod. The
p ecise cause o delayed-onse pa esis emains un e-
sol ed, al hough he ischemia- epe usion o he
spinal co d may ha e a ole. Some animals killed
immedia ely a e he ao ic unclamping showed
some conges ed a eas and essels wi h abundan ed
blood cells. This could be in e p e ed as a hype emic
phenomenon a e he ischemic insul . P obably he
epidu al cooling should combine o he s a egies
(e.g., ee adical sca enge and neu o ansmi e
an agonis ) 3'1~ o a oiding pa aplegia comple ely.
Regional hypo he mia is a common me hod o
p olonging he ischemic ole ance o many issues by
educ ion o he me abolic a e (and hus oxygen
demand). Howe e , Allen e al.2s showed ha he use
o mode a e spinal co d hypo he mia slows he
consump ion o ene gy subs a e bu does no
p e en anae obic me abolism. Al hough he op imal
Fig. 2. Radiog aph aken a e 1 minu e o epidu al
pe usion shows hea and bo h kidneys wi h adiog aphic
con as agen . Vascula up ake occu s apidly in o epidu al
eins (Ba son's plexus).
deg ee o spinal co d cooling has no been de e -
mined, some au ho s ecommend deep hypo he mia
o p o ide comple e spinal co d p o ec ion. The
e ec i eness o deep hypo he mia was shown by
Be gue e al.24 and was ecen ly con i med by
Wisselink e al.,zs who used a suba achnoid cold
luid-pe usion sys em. They demons a ed ha spi-
nal co d neu ons can ole a e empe a u es o 15 ~ o
17 ~ C wi h no his ologically iden i iable damage a e
ewa ming o 37 ~ C. Vanicky e al. 2s achie ed, by an
epidu al cooling sys em, hypo he mia below 15 ~ C.
This was con i med by us in p elimina y s udies, in
which he in used cooling luid was allowed o low
ex aco po ally. Howe e , hese echniques emain
augh wi h po en ial complica ions. An impo an
limi a ion ha we ha e encoun e ed wi h he ech-
nique o epidu al cooling desc ibed in his in es iga-
ion has been he di icul y o con olling he
in a hecal p essu e. Some animals had inc eased
in ac anial p essu es ( ou in phase I, h ee in he
me hylene g oup, and nine in he g oup gi en

JOURNAL OF VASCULAR SURGERY
Volume 23, Numbe 3
Gonzalez-Faja do e al.
451
con as ma e ial) ha disappea ed apidly a e
se e al minu es when he epidu al pe usion was
s opped. Because o he clinical use o epidu al
pe usion on humans, he high incidence o in ac a-
nial hype ension eco ded in his s udy should be
b ough o e e yone's a en ion. Al hough Da ison
e al. 31 ha e desc ibed ha CSF p essu e can be
con olled by epidu al luid d ainage, his is no
possible because epidu al a collapses he epidu al
ca he e du ing suc ion. Because in ac anial hype -
ension is due o an inc ease o CSF p essu e,
ob iously he CSF d ainage would elimina e his
complica ion.
To unde s and he kine ic p ope ies o epidu al
space, we e alua e he concen a ion- ime p o iles o
a con inuous epidu al pe usion wi h me hylene blue.
This s udy demons a es ha he spinal canal is a
highly complian space: la ge changes in olume a e
accompanied by only small changes in p essu e. The
plasma concen a ion cu e wi h a apid and linea
abso p ion implies ha he main up ake is by he
ascula compa men . Because he p ima y ac o
go e ning he abso p ion a e is local blood low, his
could explain he wide anges o olume in used in
he animals and ha he a e o pe usion has mo e
in luence on he isk o in ac anial hype ension han
he olume adminis e ed. Indeed, wi h he same
doses (10 ml adiog aphic con as agen ) i e
animals died, whe eas ou o he animals we e ali e
acco ding o he speed o pe usion. In addi ion, he
epidu al abso p ion may be pa icula ly sensi i e o
changes in osmola i y. Gi en ha he di usion o a
solu ion owa d he ascula compa men is bigge
when he pe usion is hypo onic o iso onic han
when i is hype onic, his could also explain he ac
ha he incidence o in ac anial hype ension was
highe in he g oup wi h adiog aphic con as
ma e ial han in he g oup wi h me hylene blue o
iced no mal saline solu ion. So, heo e ically, he bes
epidu al pe usion would be a hypo onic iced saline
solu ion.
On he o he hand, he co e empe a u e mea-
su ed in he ec um dec eased only a ound 2 ~ C. Thus
i appea s ha epidu al cooling is a local phenom-
enon lacking sys emic epe cussion. The adiog aphs
and spinal co d nec opsy a he end o he expe imen
showed ha he epidu al pe usion end o sp ead
mo e in a cephalic han caudal di ec ion. In addi ion,
a e epidu al pe usion, we obse ed ha he spinal
oo s and co d appea ed dyed o me hylene blue. This
ac indica es ha iced solu ion can pene a e he
du a, al hough he amoun ha .en e s he suba ach-
noid space is p obably small because o compe ing
up ake in o epidu al a and apid sys emic abso p-
ion. The iced solu ion will c oss biologic memb anes
ela i ely slowly by passi e di usion h ough he
du a. Because CSF was gene ally collec ed om a
single si e in he cis e na magna, and because he
dis ibu ion wi hin CSF is inhomogeneous, de ailed
kine ic e alua ion was impossible.
Despi e he p o ec i e e ec agains he ischemic
spinal co d damage shown in his wo k by epidu al
cooling, we mus say ha we ha e no pe o med
epidu al no mo he mic in usion in any g oup o
abbi s o compa e g oups, so i could pa ially
obscu e expe imen al cla i y. Howe e , in pilo
expe imen s designed o ma ch he expe imen al
condi ions, he epidu al ca he e inse ion and in u-
sion o no mo he mic solu ion did no p o ide any
p o ec i e e ec . Simila ly, i should be men ioned
ha he high incidence o in ac anial hype ension in
ou s udy is based on clinical signs (exoph halmos),
and no on quan i a i e objec i e da a, because
un o una ely he ele a ed CSF p essu e could be
documen ed in only one animal.
In iew o hese da a, we sugges ha he epidu al
ca he e in human beings should be placed in o he
T12-L1 in e space and ad anced only 5 cm, because
o he longi udinal sp ead o he hypo he mic solu-
ion. The placemen o a suba achnoid ca he e
becomes impe a i e o a oid neu ologic epe cus-
sions (in ac anial hype ension). This ca he e will
allow CSF d ainage and lowe CSF p essu e. The
educ ion o CSF p essu e will pe mi an inc ease o
passi e di usion o iced solu ion in o he suba ach-
noid space. Consequen ly, he epidu al pe usion
sys em o egional co d hypo he mia will be sa e
and mo e e ec i e in lowe ing CSF empe a u e. We
belie e he echnique desc ibed by Da ison e al. 31 is
simple and e ec i e in lowe ing CSF empe a u e
and will p o ide p o ec ion om ischemic spinal co d
inju y du ing he c i ical pe iod o ao ic c oss-
clamping and in e cos al essel eanas omosis. How-
e e , his me hod p obably does no a oid he isk o
delayed-onse pa aplegia, as i has been showed in ou
s udy. In addi ion, he osmola i y and a e o epidu al
pe usion a e he mos impo an ac o s in he
de elopmen o in ac anial hype ension. Fu he
in es iga ions a e necessa y o explain hese obse a-
ions and he possibili y o epe usion inju y.
REFERENCES
1. Shenaq SA, S ensson LG. Pa aplegia ollowing ao ic su ge y.
J Ca dio ho ac Vase Anaes h 1993;7:81-94.
2. S ensson LGG, Von Ri e CM, G oene eld HT, e al.
C oss-clamping o he ho acic ao a: in luence o ao ic
shun s, laminec omy, papa e ine, calcium channel blocke ,
JOURNAL OF VASCULAR SURGERY
452
Gonzalez-Faja do e al.
Ma ch 1996
aUopu inol, and supe oxide dismu ase on spinal co d blood
low and pa aplegia in baboons. Ann Su g 1986;204:38-47.
3. Agee JM, Flanagan T, Blackbou ne LH, K on IL, T ibble CG.
Reducing pos ischemic pa aplegia using conjuga ed supe ox-
ide dismu ase. Ann Tho ac Su g 1991;51:911-5.
4. Bowe TC, Mu ay MJ, Glo iczki P, Yaksh TL, Hollie LH,
Pai ole o PC. E ec s o ho acic ao ic occlusion and ce e-
b ospinal luid d ainage on egional spinal co d blood low in
dogs: co ela ion wi h neu ologic ou come. J VASC SuR~
1988;9:135-44.
5. Qayumi AK, Janusz MT, Jamieson WRE, Lys e DM.
Pha macologic in e en ions o p e en ion o spinal co d
inju y caused by ao ic c ossclamping. J Tho ac Ca dio asc
Su g 1992;104:256-61.
6. Elmo e JR, Glo iczki P, Ha pe M, e al. Spinal co d inju y in
expe imen al ho acic ao ic occlusion: in es iga ion o com-
bined me hods o p o ec ion. J VASC SuRe 1992;15:789-99.
7. C aw o d ES, S ensson LG, Hess KR, e al. A p ospec i e
andomized s udy o ce eb ospinal luid d ainage o p e en
pa aplegia a e high- isk su ge y on he ho acoabdominal
ao a. J VASC SURG 1990;13:36-46.
8. S ensson LG, G um DF, Bedna ski M, Cosg o e DM III,
Loop FD. App aisal o ce eb ospinal luid al e a ions du ing
ao ic su ge y wi h in a hecal papa e ine adminis a ion and
ce eb ospinal luid d ainage. J VASC S Rc 1990;11:423-9.
9. Ache CW, Wynn MM, Hoch JR, Popic P, A chibald J, Tu -
nipseed WD. Combined use o ce eb al spinal luid d ainage
and naloxone educes he isk o pa aplegia in ho acoabdomi-
nal aneu ysm epai . J VASC SuRe 1994;19:236-48.
10. Wisselink W, Money SR, C ocke DE, e al. Ischemia-
epe usion inju y o he spinal co d: p o ec i e e ec o he
hyd oxyl adical sca enge dime hyl hiou ea. J VAsc SuRc
1994;20:444-50.
11. Ki shne DI, Ki shne RI, Heggeness LM, DeWeese JA.
Spinal co d ischemia: an e alua ion o pha macologic agen s
in minimizing pa aplegia a e ao ic occlusion. J VASC SuRe
1989;9:305-8.
12. F ancel PC, Long BA, Malik JM, T ibble C, Jane JA, K on IL.
Limi ing ischemic spinal co d inju y using a ee adical
sca enge 21-aminos e oid and/o ce eb ospinal luid d ain-
age. J Neu osu g 1993;79:742-51.
13. Pon ius RG, B ock nan HL, Ha dy EG, Cooley DA, De-
Bakey ME. The use o hypo he mia in he p e en ion o
pa aplegia ollowing empo a y ao ic occlusion: expe imen-
al obse a ions. Su ge y 1954;34:33-8.
14. S ensson LG, C aw o d ES, Pa el V, McLean TR, Jones JW,
DeBakey ME. Spinal oxygena ion, blood supply localiza ion,
cooling, and unc ion wi h ao ic clamping. Ann Tho ac Su g
1992;54:74-9.
15. Naslund TC, Hollie LH, Money SR, Facundus EC, Skend-
e is BS. P o ec ing he ischemic spinal co d du ing ao ic
clamping: he in luence o anes he ics and hypo he mia. Ann
Su g 1992;215:409-16.
16. Robe son CS, Fol z R, G ossman RG, Goodman JC.
P o ec ion agains expe imen al ischemic spinal co d inju y.
J Neu osu g 1986;64:633-42.
17. Rokkas CK, Sunda esan S, Shuman TA, e al. A p ima e
model o spinal co d ischemia: e alua ion o spinal co d blood
low and he p o ec i e e ec o hypo he mia. Su g Fo um
1991;42:265-7.
18. Hollie LH, Money SR, Naslund T, e al. Risk o spinal co d
dys unc ion in pa ien s unde going ho acoabdominal ao ic
eplacemen . Am J Su g 1992;164:210-4.
19. Kouchoukos NT, Wa eing TH, Izumo o H, Klausing W,
Abboud N. Elec i e hypo he mic ca diopulmona y bypass
and ci cula o y a es o spinal co d p o ec ion du ing
ope a ions on he ho acoabdominal ao a. J Tho ac Ca dio-
asc Su g 1990;99:659-64.
20. C aw o d ES, Coselli JS, Sail HJ. Pa ial ca diopulmona y
bypass, hypo he mic ci cula o y a es , and pos e ola e al
exposu e o ho acic ao ic aneu ysm ope a ion. J Tho ac
Ca dio asc Su g 1987;94:824-7.
21. Colon R, F azie OH, Cooley DA, McAllis e HA. Hypo-
he mic egional pe usion o p o ec ion o he spinal co d
du ing pe iods o ischemia. Ann Tho ac Su g 1987;43:639-
43.
22. Coles JG, Wilson GJ, Sima AF, e al. In aope a i e manage-
men o ho acic ao ic aneu ysm. J Tho ac Ca dio asc Su g
1983;85:292-9.
23. Allen BT, Da is CG, Osbo ne D, Ka l I. Spinal co d ischemia
and epe usion me abolism: he e ec o hypo he mia. J V^sc
S ;Re 1994;19:332-40.
24. Be gue R, Po o J, Fedo onko B, D ago ic L. Selec i e deep
hypo he mia o he spinal co d p e en s pa aplegia a e ao ic
c oss-clamping in he dog. J VAsc SURG 1992;15:62-72.
25. Wisselink W, Becke MO, Nguyen JH, Money SR, Hollie
LH. P o ec ing he ischemic spinal co d du ing ao ic
clamping: he in luence o selec i e hypo he mia and spinal
co d pe usion p essu e. J VASe S ;Rc 1994;19:788-96.
26. Neg in J J , Klaube LD. Di ec egional hypo he mia o he
cen al ne ous sys em: a p elimina y epo o a pilo p ojec
on expe imen al hypo he mia. A ch Neu ol 1960;3:100-4.
27. Ma sala M, Vanicky I, Galik J, Radonak J, Kund a I, Ma sala
J. Panmyelic epidu al cooling p o ec s agains ischemic spinal
co d damage. J Su g Res 1993;55:21-31.
28. Vanicky I, Ma sala M, Galik J, Ma sala J. Epidu al pe usion
cooling p o ec ion agains p o ac ed spinal co d ischemia in
abbi s. J Neu osu g 1993;79:736-41.
29. Kaschne AG, Sandmann W, Kniemeye HW, Schie R,
La kamp H. E alua ion o epidu al pe usion cooling o
p o ec he spinal co d du ing ho acic ao ic c oss-clamping:
moni o ing o spinal e oked elec og am. J Ca dio asc Su g
1985;26:97-8.
30. Tabayashi K, Niibo i K, Konno H, Moh i H. P o ec ion om
pos ischemic spinal co d inju y by pe usion cooling o he
epidu al space. Ann Tho ac Su g 1993;56:494-8.
31. Da ison JK, Camb ia RP, Vie a DJ, Columbia MA, Kous as
G. Epidu al cooling o egional spinal co d hypo he mia
du ing ho acoabdominal aneu ysm epai . J VAsc SuRe
1994;20:304-10.
32. Ta lo IM. Spinal co d comp ession: mechanisms o pa alysis
and ea men . Sp ing ield, Ill.: Cha les C. Thomas, 1957:1-
147.
33. Roby JF, Whi e BJ. Biochemical echniques: heo y and
p ac ice. Mon e ey: B ooks/cole Publishing, 1987:40-72.
34. Zi in JA, de Gi olami U. Spinal co d in a c ion: a highly
ep oducible s oke model. S oke 1980;11:200-2.
35. Moo e WM, Hollie LH. The in luence o se e i y o spinal
co d ischemia in he e iology o delayed-onse pa aplegia. Ann
Su g 1991;213:427-32.
Submi ed Ma ch 23, 1995; accep ed July 25, 1995.