BASIC SCIENCE SECTION

onosialoganglioside G 1 treatment after a hypoxic-ischemic episode reduces the vulnerabili

fetal sheep brain to subsequent injuries

William K.M. Tan, MSc, Chris E. Williams, PhD, Carina E. Mallard, BSc, and Peter D. Gluckman, DSc

Auckland. New Zealand

OBJECTIVE: A transient hypoxic-ischemic episode can markedly sensitize regions of the fetal brain, particularly the striatum, to further insults occurring in close succession. The purpose of this study was to determine whether ganglioside GM1 given to the fetus after a hypoxic-ischemic episode can counteract this sensitization and protect against subsequent insults. STUDY DESIGN: Chronically instrumented near-term fetal sheep (119 to 133 days) were subjected to three 10-minute episodes of reversible cerebral ischemia at 1-hour intervals. Six were given 30 mg/kg ganglioside GMl through the umbilical vein. commencing immediately after the release of the occluder, over the next 2 hours, followed by a continuous infusion of 30 mg/kgJday over 60 hours after ischemia; these were compared with seven vehicle-treated controls. The time course of electrocorticographic activity and cytotoxic edema within the parasagittal cortex were determined with real-time spectral analysis and continuous impedance measurements, respectively. The degree and distribution of neuronal death was assessed histologically 72 hours later. RESULTS: Ganglioside GM1 improved recovery of primary edema and markedly reduced histologic damage CP < 0.001), particularly in the striatum, hippocampus, and cortex. At 72 hours after ischemia electrocorticographic activity had returned to normal in the ganglioside GM1-treated group but was still depressed CP < 0.001) in the control group. CONCLUSION: These results showed that ganglioside GM1 treatment initiated immediately after a transient hypoxic-ischemic episode stabilized membrane function and markedly improved neuronal outcome after subsequent insults. suggesting its potential therapeutic value in acute situations of repeated hypoxia-ischemia in the perinatal period. (AM J OSSIEI GYNECOL 1994;170:663-70.)

Key words: Ganglioside GM\. repeated cerebral ischemia, fetal sheep

of the

Most perinatal asphyxial insults are thought to occur as antepartum or intrapartum events, I and cerebral damage is thought to be a consequence of asphyxia associated with a period of impaired cerebral blood flow or ischemia.2 Experimental studies have shown that brief repeated hypoxic-ischemic injuries interact syner­gistically and worsen outcome compared with isolated episodes in animals.'

A number of strategies have been suggested to pro-

From the Research Centre for Drotlopmtmal Medicine and Biology, School of MedICIne, University of Auckland.

teet the central nervous system against hypoxic-isch­emic injury. However. few experimental studies have considered the effect of therapeutic interventions on injury induced by repeated brief episodes of hypoxia­ischemia. Given that in most situations of perinatal asphyxia therapy will only be possible after the first recognized asphyxial episode, strategies that reduce secondary neuronal death may be particularly impor­tant. Such neuronal death is, in part, a result of the repeated asphyxial insults and a consequence of de­layed neuronal loss secondary to mechanisms initiated during the primary insult.·

Supported by grants from the Health Research Council of New Zealand, the Nturologual FoundatIon, and the NatIonal Children's Health Research FoundatIOn of New Zealand. W. T. was supported by Tan Kah Ktt FoundatIon Postgraduate Scholarship, Smgapore. RtctlvtdforpubllcationApril16, 1993; Ttllised Septtmber8, 1993; acupted Sl'pttJnber 29, 1993. Rtpnnts not available. Copyright © 1994 by Mosby-Year Book, 1nc. 0002-9378194 $3.00 + 0 611151816

Monosialogangliosides GM\ are sialylated glyco­sphingolipids that have been found in many tissues but with the highest concentration in the nervous system. ~ The ganglioside molecule is an important component of cell membranes, with the ceramide incorporated into the phospholipid layer and the oligosaccharide chain with sialic acid on the membrane surface.b When ad-

663

664 . Tan et al.

ministered intra peritoneally, GMt appears to cross the placentaF and blood-brain ball iers· and is incorpo­rated into the neural cell membranes. 9 Exogenous GMt has been shown in adult animals to protect acute processes of ischemic injury, thereby enhancing subse­quent structural and functional recovery. to Our previ­ous data show that treatment with GMt commencing before an insult can protect the fetal sheep brain against a single severe hypoxic-ischemic episode with­out adverse physiologic and metabolic effects. II

We recently reported that in a chronically instru­mented near-term fetal sheep repeated brief episodes of reversible cerebral ischemia altered the distribution of and aggravated the degree of neuronal loss.' The striatum was particularly sensitized, inducing an en­cephalopathy similar to that observed in some asphyx­iated term infants in whom choreoathethoid cerebral palsy later developed. '2 The purpose of this study was to determine whether GM t administered to the fetus after the first episode of hypoxia-ischemia can counter­act this sensitization and protect the fetal brain against subsequent insults.

Material and methods Surgical procedures. Sixteen Romney/Suffolk near­

term fetal sheep from 119 to 133 days of gestation were operated on under halothane anesthesia (2%) by means of sterile techniques, as previously described." Briefly, the head, neck, and forelimbs of the fetus were exter­nalized, and catheters were inserted into the brachial arteries, umbilical vein, brachial vein, and amniotic cavity. Three pairs of shielded stainless steel electrodes were placed over the parietal dura, two at 1 0 mm lateral to the bregma and 5 mm and 15 mm anterior, and the third at 15 mm lateral and 10 mm anterior to bregma. A pair of electrodes was also sewn into the paraspinal nuchal muscle to record e1ectromyographic activity. The vertebral-occipital anastomoses between the ca­rotid arteries and vertebral arteries were ligated bilat­erally to eliminate vertebral blood supply to the brain. Inflatable occluder cuffs were placed around both ca­rotid arteries. The fetus was returned to the uterus, and the uterine and abdominal walls were closed. After the operation the ewe was housed in a metabolic cage at a constant temperature (200 C) and humidity (50%) and given free access to hay and water, supplemented by sheep nuts and alfalfa. Antibiotics (gentamicin, 80 mg; penicillin, 500 mg) were administered to the ewe daily.

Recordings. A four-electrode technique was used to measure changes in impedance associated with changes in extracellular space that occur concomitantly with cytotoxic edema." The impedance signal, electrocor­ticogram, and nuchal electromyogram were recorded on an analog chart running at 5 mm/min. Electrocor­ticographic intensity spectra were analyzed on-line as previously described."

•

February 1994 Am J Obstet Gynecol

Experimental procedures. Experiments were per­fOllued 72 hours after surgery. Fetal arterial samples were obtained before the start of each experiment, and only fetuses with normal arterial blood gases (pH > 7.32 and Pao2 > 17 mm Hg) were used. After a baseline recording period of 12 hours the carotid cuffs were inflated with saline solution for 10 minutes each time; the inflation was repeated 3 times at I-hour intervals. Successful occlusion was confirmed by an isoelectric electrocorticogram. Three animals were re­jected from the study after the first occlusion because of incomplete suppression of the electrocorticogram dur­ing the occlusion, presumably reflecting inadequate ligation of the vertebral-occipital anastomoses. The re­maining animals were assigned to two groups, with one group receiving GM I (Fidia Research, Abano Terme, Italy) (n = 6) and the control group (n = 7) receiving equivalent volumes of phosphate buffer infusion only. In the former group 30 mglkg of GM I dissolved in phosphate buffer (pH 7.4) was administered systemi­cally for 2 hours through the umbilical vein immedi­ately at the end of the first 10-minute ischemic insult, followed by a continuous infusion of 30 mglkglday administered at a rate of 1.2 ml/hr by means of a syringe pump (Harvard Apparatus, Millis, Mass.) over 60 hours after ischemia. These experiments were ap­proved by the Animal Ethics Committee of the Univer­sity of Auckland.

Histologic examination. Each sheep was killed by pentobarbital injection 3 days after the ischemia. The fetal brain was immediately perfused through the ca­rotid arteries with 500 ml of physiologic saline solution, followed by 500 ml of 10% formalin. The fixed brain was removed and placed in 10% fOllualin for 2: 24 hours. After processing and wax embedding in paraffin, coronal subserial sections were cut 8 (.Lm thick and then stained with thionin-acid fuchsin. Every fortieth section was examined by light microscopy by an independent assessor who was blinded to the experiment. Neurons with ischemic cell change, consisting of acidophilic (red) cytoplasm and contracted nuclei or with just a thin rim of red cytoplasm with pyknotic nuclei, were assessed as dead, whereas all others were considered viable. Each region was scored for the proportion of dead neurons, as described previously."

Analysis. The changes in electrocorticographic in­tensity were transformed to a log ratio scale (db)" and normalized with respect to the 12-hour reference pe­riod before occlusion. A digital Blackman low-pass filter with a cutoff of 0.1 cycles/point was applied to the electrocorticographic intensity data to minimize short­term fluctuations of < 20 minutes. Cortical hyperexcit­ability was defined as electroencephalographic intensity 2: 5 db above the reference period. '4 We have previ­ously shown that this pattern of hyperexcitability is associated with e1ectromyographic evidence of seizure

Volume 170. Number 2 Am J Obstet Gynecol

l' iI'.I'

Tan at al. 665

311 AFTER ISCHEMIA

_'.'11 ~

A

ECOG 1 __ ~j L...! __ ---l' L __ u I

• • 211 GM 1 INFUSION 11011 GM 1 INFUSION

(30 mg/kg) 130 mg kg d8Y)

B

[COG • i 4 ,', .. 2 ~Ji.'r £ .1 ". -! .. • ... ¢i~/I ... ! ....... _ ... : ; __ ....

I i

3rd·occluslon L i I

lal occlu.!on 2nd occlua!on 10 MINS • •

211 VEHICLE INFUSION 6011 VEHICLE INFUSION

Fig. 1. Representative lhart recording~ showing changes in raw elcnro<Orticographic (F-COG) activity in response to three IO-minute occlusions repeated I hour apart in GMI-treated animal (A) and vehicle-treated control. In A. 30 mg/kg GM, wa~ administered ~y~temilally for 2 hours through umbilical vein immediately after end of fir,t 10-minute ischemic IOsult. followed by continuous infusion of GM, (30 mg/kg/day) over 60 hour~ after i~chemia. In B. colltrol fetmes received equivalent volumes of phosphate buffer infusion only. At 3 hours after IS( hernia t. M ,-treated fetme, showed greater recovery of electrocorticographic aoivity than did vehicle-treated control,.

Table I. Changes in fetal blood gas and metabolic status in response to three 10-minute episodes of

ischemia 1 hour apart at 72 hours after ischemia

Vthlrie-trealed rolllrol.1 GM rtTfated jelll.les (n = 7. gel/a/1OMI flge 127 ± 1 day~) (n = 6. geltatlOnal flf(e 126 ± 1 tin.rs)

Pao? Paco? - • pH (mm Hg) (mill Hg)

Before first 7.38 ± 0.01 23 ± 1 47 ± I ischemia

Before second 7.33 ± 0.03 25 ± I 48 ± 3

ischemia Before third 7.28 ± 0.04* 23 :t: I 51 ± 2

ischemia At 72 hr 7.36 ± 0.02 25 ± 2 48 ± 2

Values are expressed as mean ± SE. *p < 0.05 compared with first ischemia. tp < 0.001 compared with first ischemia.

!P < 0.05 compared with vehicle control~.

Lactate (mmol!L)

0.9 ± 0.1

2.8 ± 1.3

4.2 ± 0.9t

1.7 ± 1.1

activity." The final electrocorticographic intensity was calculated as the average of the final 4-hour period. from 68 to 72 hours. Changes in cerebral impedance were expressed as the percentage of preinsult levels."

The time course of electrocorticographic intensity and impedance in the two groups and changes in histologic outcome were compared with two-way analy­sis of variance with time or regions as repeated mea­sures where appropriate. followed by the Newman­Keuls test for multiple comparisons. All results are

presented as mean ± SEM.

Results The gestational age and physiologic index values

before insult were similar for the GMt -treated and vehicle control fetuses (Table I). The body weight of GM t -treated fetuses (3.7 ± 0.3 kg) did not differ from that of vehicle controls (3.8 ± 0.2 kg).

Glu({).Ie PflO, Pfl(,O, Lactate Gluco.le • • (mmollL) pH (mm Hg) (mm Hg) (mmol!L) (mmo/IL)

0.8 ± 0.1 7.38 ± 0.01 23 ± 1 47 ± I 1.0 ± 0.2 0.8 ± O. I

1.0 ± 0.1 7.37 ± 0.01 26 ± I 47 ± I 1.6 ± 0.4 1.0 ± 0.1

1.0 ± 0.1 7.37 ± O.O]! 24 :t: ] 45 :t: 2 1.8 + O.5t o.g ± 0.1

0.7 + 0.1 7.36 ± 0.01 22 ± I 48 ± 2 1.0 ± 0.3 0.8 ± 0.1

During each bilateral carotid occlusion the electro­corticogram became isoelectric. After the release of the occluder cuffs the electrocorticogram progressively in­creased in intensity in both groups of animals (Figs. 1 and 2). The GM I group showed greater electrocortico­graphic activity during the first 12 hours after ischemia ( - 1.89 ± 1.02 db with respect to baseline vs -6.99 ± 2.46 db.p < 0.(01) (Fig. 3). During the next 48 hours the intensity of the electrocorticogram was greater in the treated than in the control group (p < 0.05). No postinsult cortical hyperexcitability was observed. except in one control animal where hyperex­citability developed 5.1 hours after the thIrd occlusion and lasted for 8.2 hours. At 72 hours after ischemia the electrocorticographic intensity of GM I-treated fetuses was similar to baseline (- 0.5 ± 0.9 db). whereas the controls had significantly reduced intensity (-7.7 ± 2.6db.p < 0.(01).

666 Tan et aI.

Ji ..

., ..

'" .... -1'1 R

o -5

! -10 R -U -15 o u -20 ~

-. .. 140 --• u 130 .. • ...

120 • 110 B

110 ... ... • u 100 -.. .. 0 u

Control GM1-treated

February 1994 Am J Obstet Gynecol

o 5 10 15 20 25 o 5 10 15 20 25

Time (hoUl~s)

Fig. 2. Time Course of changes in electrocorticographic (ECoG) intensity (uppn) and cortical impedance (lower) in representative GMI-treated fetal sheep (nght) and control animal (left) after three 10-minute episodes of reversible cerebral ischemia. Time is shown as hours after first ischemic insult was induced.

... .. ~ "C

o ., t' -2

~ -4 ~ .... rn r:: -6 CI) ~ r:: -8 • I

t:J -10 o

U tIJ-12

Time (hours)

0-12 12-14 U-38 38-48 48-80

... ... ... ... ...

... ... ...

Fig. 3. Bar graph comparing time course of changes in mean electrocorticographic (ECoG) intensity after three lO-minute episodes of reversible cerebral ischemia in GM I-treated fetuses (c/t'ar) and vehicle controls (fdled) Time is shown as hour~ after ischemia. Values are means :t SE. Allntlk, P < 0.05; two altfT­

ISks, p < 0.0 I, thrtt a.5trruks, p < 0.00 I.

Cortical impedance rapidly increased from 5 ± 2 minutes after occlusion and peaked at 3 ± 2 minutes after release of the clamps (Fig. 2). The cortical imped­ance increased during each occlusion similarly in both groups. After recirculation the impedance did not fully recover to baseline levels before the second and third

occlusions but partially resolved with some residual increase in impedance. The impedance after the first and second occlusions in GM\-treated fetuses (100.4% ± 0.6% and 101.1% ± 0.7%) was not signifi­cantly different from that of vehicle controls (100.8% ± 0.6% and 104.5% ± 1.6%). At 3 hours after the third occlusion the residual impedance was signifi­cantly greater in the controls (109.7% ± 4.7%) than in the treatment group (100.7% ± 0.7%, P < 0.01) (Fig. 4). In the treated animals impedance remained low, with no evidence of a secondary rise. However, in the control group, although there were large variations, there was evidence of persistent cytotoxic edema (p < 0.05). At 72 hours after ischemia the residual impedance was lower in treated animals (102% ± 2.1 %)

than in the controls (117.5% ± 17.4%), although this difference did not reach statistical significance.

A progressive metabolic acidosis developed with each occlusion in both groups (fable I). Lactate levels in­creased after each successive occlusion but to a greater degree in the controls. No significant changes were observed in Pao2 during the experiment. At 72 hours blood gases and metabolic status between the two groups were not significantly different (fable I).

On histologic evaluation the ove. all reduction in neuronal loss in the GM I-treated animals was highly significant (p < 0.01) (Fig. 5). In the control animals there was > 30% neuronal damage in the parasagittal

Volume 170, Number 2 Am J Obstet Gynecol

Cortical impedance " Pre-ischemia

120

115

110

105

100 0-1

Cortical impedance " Pre-hchemia

150

140

130

120

110

100

• • • •

1-2 2-3 3-4 4-5 5-11

* 11-12 12-24 24-311 311-411 48-110

Time (hours)

Fig. 4. Bar graphs comparing mean conical impedance dur­ing first 6 hours (upper) and from 6 to 60 hours (lower) after three IO-minute occlusions repeated 1 hour apan in GM t-treated fetuses (clear) and vehicle-controls (filled). Values are means ± SE. Asterisk, P < 0.05; two astn'isks, p < 0.01; three asttlisks, p < 0.001.

cortex, striatum, cornu amnionis 1, 2, 3, 4, and amygdala (Figs. 6 through 8). In contrast, in no regions of the GMJ-treated animals was there any significant neuronal loss. The parasagittaI cortex, striatum and cornu amnionis I, 2, 3, and 4 subfields of the hippo­campus exhibited the greatest level of protection, >80%.

Comment

This study was designed to mimic a potentially clin­ically relevant paradigm in that GM 1 was administered after the first episode of fetal asphyxia. Our results demonstrated that the increased central nervous system vulnerability to injury induced by multiple episodes of brief cerebral ischemia was reversed by GM) treatment commencing immediately after the first episode. as indicated by reduced neuronal loss particularly in the striatum, hippocampus, and cortex and no loss of electrocorticographic activity at 72 hours after injury.

PSCX

LT J.

STR

DG CA4

CA3 CA1,2 THA

AMG

I

•••

• ••

••

Tan et al. 667

Control

DCIU

I I I I I

o 20 40 60 80 100

NEURONAL LOSS SCORE

Fig. 5. Histologic demonstration of neuronal damage in var­ious brain regions at 72 hours after three lO-minute episodes of reversible cerebral ischemia at I-hour intervals. Damage scores are on linearized scale, 0 to 100 (0, no neuronal loss; 100, total necrosis). Values are means ± SE. GMt-treated animals showed significantly less damage (two astensks, p < 0.01; three astmsks, p < 0.001). PSCX, Parasagittal conex; LTCX, lateral conex; STR, striatum; DG, dentate gyrus; CAl, 2, J, 4, cornu amnionis 1. 2, 3, and 4 of the hippocampus; THAL, thalamus; AMG, amygdala.

The improved outcome after GM l therapy is consistent with our previous report in fetal sheep" and the reports of others in adult animals demonstrating that systemic GM 1 treatment can reduce morphologic, biochemical, neurophysiologic, and behavioral manifestations of hypoxic-ischemic brain damage. If)

A number of mechanisms have been implicated in hypoxic-ischemic cell loss, including intracellular so­dium and water accumulation leading to lysis, intracel­lular calcium accumulation, and toxic effects of excita­tory amino acids. I ~ The mechanism of action of G M J is not fully understood. Given the effectiveness of GM J in a broad spectrum of central nervous system injury paradigms, its mechanism of action may involve an effect on some cellular processes common to injury in both developing and mature central nervous system tissue. It has been suggested that GM J is incorporated into cell membranes and that this results in a stabiliza­tion of membrane integrity and function.6

On-line impedance is a technique to estimate changes in extracellular space that occur concomitantly with cytotoxic edema and reflect cell membrane func­tion. It, The residual impedance seen after the repetitive insults in the control group indicates that membrane function was impaired. 16 I n contrast, G M I -treated fe-

668 Tan et al.

a

, ,

• , . , .-­.. . . , • • . ' ,

, , . . .. ,~

0' ':~. .. '" .. " I '" .,. -1". • .,

o " " ~ " ', .. ,, 'toO, .' .,.. ~., . ,

• , • . 1 " I. 'I · .( _ ... ", .. p-' ..

~ I,~" '. , ~

,~ '~1

~.". "r ' ·~ "· ... .. ... t . ..

. , , , '" . t"

~ ; • • .. · ·A·4:" ': 1'-., &..i ~. • ~ ~ ." • ".* 11",. , ' •• j~'.ftJ~ • •• :' ., . .... ,t.

It • ., • . • I 1 ... .. f t ..

~ , .,~4". 1 .. ~~ .. ." l''' j'/.';·f ~. ~~. c" r" ,... ... ".

, "'>( I ..- t~. t ~ *,' :---,+ 1'

~ •• II

,

t

,

. " c

I. '1.).04 • I' '., " • ..... ..

"J .\. ••• ~~.:~ .. " ..... "-• 't ~ - . I··.. "'I.~ .. . , . . ,.. I • • ...... ,. • , • , 'a ••

I , • \ .. . . ".. ... ":~'\,. '.~.4' ."

;' .'~, .; .... ;, .... ' ,." "

' • • I~ .. It,/ ~ • , ", . ~' jl' 'l n.:. t) '.t' .-'J'" 4 f.. ~.

" , .' .. ~, 1··· .. f4.~J ',"J , I ,#~ I' ...-"

I • •• ~...t" 1" ~, ~ I •• J'fi I( .... ~., , It ':., i,", I /,,/., r;

• ~l' • ~ .. "I '.'. , f ./ I " . , , fI;" • i =I , • ., ... t" "'~. ~"';,. ~ , . .1'.,.'. ""I", • -".f :'.- t .••. I; t · ~. <If· ,- #S- I ... ~.. -'''-..... . • t . .. t\ • • v .,I.- i ,i>. F ,. ..; • ~ ., ~.~ •• •

, ,........~ I,' -" t, • • " -* ,J i • .,.· ...... .. •. -1 "..-"".. ; ,,~ • ., 4 I-to! .......

, ', t 'I" J" •• ' :' I·;" t f.. .. ,~, ... or; ~ .. • ,. •

,.ttl . '.' ~ -, , 11'" ' ..... '.! .. . - . .. , ,

• . ,

.... "t- , , - -

, •

V r

. ' o

"

6 *. • •• ..- . , . "

'd .. • •

• • •

Febru"r, 199-1 Am J Obslel G. nceol

,

Fig. 6, Reprc~entalive phowmi rograph of para~agillal region of parietal corlex at 72 hours after 10 minutes of cerebral i <-hernia repeated three tim s at I-hour inter\'al~ in fetal sheep, a, b, Vehi Ie-control animals; c, d, G {, treatment (30 mg/kg) inilial d immediat ly afler flrt isch mi in sult. Pyramidal cells (within melt) are shown under high magnifkation (b, d), Bar, 130 IJ.m (a, c) and 30 IJ.m (b, d) . r rhionin- arid f ch~in,)

tu e~ ~howed I s resiollal cyrotoxi edema shortly after ischemia. This is con i tent with previoLis observations" and further support the hy that GM I treat­ment protects membrane function during th acute phase of th i chemic ir~jury < nd reperfu~ion and in­crca e the potential for recovery of the injured neuron .'7

GM I Gin protect neurom against excitotoxicity. tuel­ies have ~h()wll that GM I altenuale~ esential amino acid-related n uroroxi it in neuronal <ell cultures.'~

\'\'h 11 aclmini~tcrcd ~y~tcmicall" LO neonatal animal. GM I reduce~ neuronal damage indut d b)' exogenous exeiroLOxim.'" I'he allliexciwtoxic el1ccts of (,M I ap-

pear to oc lIr by inhibition of intracellular c\'ents. M I has be n reported to inhibit glutamate-m dial d tran -10 arion of protein kina. e from cyto 01 to cell mem­brane, thereby limiting the incrca ' in intrac lIular fr Ca " in primary granular cereb Ilar cultures"" and in fetal rat br< in,7

"'1' 1, has also b en hO\\ln to potentiate the action of a large numb r of 11 uronotrophic factor, in luciing nerve grm th faClor, both in \ io'o and in \ ivo.11 Pre 'urn­ably by a<.ling on pla~m" ll1embran s to ameliorate the a ute ~tage~ of injUl), G \1 I prm ide, fa\l)1 able ndi­tions to allow a r spome 10 trophil fa(l()r(~) .

The comept of in utero nellroproteuion has not

Volume 170, Number 2 Am J ObSLCL Gynccol

I

!v

• • • • •

• •

.. •

J

Tan at al. 669

• • ..

f

•

• Fig. 7. Photomicrographs howing striata l neuron 72 hour alter 10 minut ~ of cerebral i, h mia repealed three time; at I-hour interval in vehicle-control (a) and GMI-treaLed (b) fetal sheep. Bar, 30 !-LIll, (Thionin-a id fu hsin.)

• • • •

o • .. L '

Fig. 8. Dor al hippocampus 72 hours after 10 minute, of cerehral i chemia repeated thre tim . at I-hour int Ivai in fetal sheep . a, h, antrol fetuse~; c, d, C,M I-treated :mimals. muum pyramidale in cornu amnionis 1 legion (bet" en a1TOlL's) i hm,n unclel high magnifi alion (b, d). Bar, 0.7 mm (a c) and 30 !-Lm (b, d) . (T 'hionin-acid fuchsin.)

be n exten. ivel con id r d. The potcmiaJ appli alion of mall} proposed treatment for i olated insult in th peri mil a! p 'riod is limited."' In the fetal ral \ubjected to

lIml iii al cord 0 elusion prophylactic matel nal admin­i tration of nimodipinc reduces postnatal hi loch mica!

derangcment and b h. ,,-jom! imp.lirlllcnt. "·\ However.

lIealmenl wilh the calcium hann I ulllagoni IS nicar­dipll1c or f1unari7in call cau~e ardia depr \ ion and death al high dose in fi tal ,h epJl or ne,\ bom in­

fants. "" The CUITem Iud), togethcl "illl ollr previolls

670 Tan et aI.

.

observations. II suggests the potential for acute thera-peutic interventions with GM. in the perinatal period. However. we have only shown that GM. treatment does not cause hypotension or metabolic perturbations in the healthy near-term fetal sheep. II Given that hy­potension and cardiovascular compromise secondary to asphyxia appear to be the primary cause of cerebral injury. further studies are required to determine the effects of GM. therapy on cardiovascular function un­der conditions of severe systemic asphyxia. such as umbilical cord occlusion. GM. may also have worth­while benefits for chronic high-risk situations; however. long-term functional outcome will need to be carefully evaluated to determine the risks and benefits of long­term administration. Administration of GM. through the umbilical vein may present practical difficulties that may delay treatment. However. a study by Magal et al. 7

suggests that GM. can cross the placenta. but the maternal-fetal kinetics are not yet known and need to be established.

In summary. we designed the study to mimic the clinical possibility of initiating neuroprotective therapy immediately after the diagnosis of fetal distress. Our observations demonstrate that ganglioside GM. given after an episode of hypoxia-ischemia protects the fetal sheep brain from the effect of cumulative insults. Given that GM. appears to cross the placental barrier7 and lacks systemic side effects II that are observed with other potential prophylactic agents such as glutamate and calcium antagonists. GM. therapy in utero is a potential strategy to protect the central nervous system of acutely distressed fetuses at risk for further hypoxic-ischemic injuries. However. caution must be exercised before extrapolating these results into a treatment modality for clinical use.

We thank Fidia Research Laboratories. Abano Terme. Italy for the gift of GM •.

REFERENCES I. Bejar R. Wozniak P. Allard M. Bernirschke K. Baucher Y.

Antenatal origin of neurologic damage in newborn infants. AM.I OOSIEI GYNEWL 1988;159:357-63.

2. Volpe.l. Herscovitch P. Perlman .1M. Kreusser K. Raichle ME. Positron emission tomography in the asphyxiated term newborn: parasagittal impailment of cerebral blood flow. Ann Neurol 1985; I 7:287-96.

3. Mallard EC, Williams CEo Gunn A.I. Gunning MI. Gluck­man PD. Frequent episodes of brief ischemia sensitize the fetal 'heep brain to neuronal loss and induce striatal injury. Pediatr Res 1993;33:61-5.

4. Gluckman PD. Williams CEo When and why do brain cells die? Dev Med Child Neurol 1992;34:1010-4.

5. Nagai Y. Iwamori M. Ganglioside distribution at different levels of organisation and its biological implications. In: Ledeen RW. Yu RK. Rapport MM. Suzuki K. eds. Gangli­OSIde structure. function. and biological potenual. New York: Plenum. 1984: 135-46.

6. Ghidoni R. Fiorilli A. Trinchera M. Venerando B. Chig­ornO V. Tettamanti G. Uptake. cell penetration and met­.. boli«. processing of exogenously administered GM I gan­glio~ide in rat brain. Neurochem Int 1989; 15:455-65.

February 1994 Am J Obslel Gynecol

7. Magal E. Louis Jc, Aguilera J. Yavin E. Gangliosides prevent ischemia-induced down-regulation of protein ki­nase C in fetal rat brain. J Neurochem 1990;55:2126-3 J.

8. Orlando p. Cocciante G. Ippolito G. Massari P. Robeni S. Tettamanti G. The fate of tritium labelled GM I ganglio­side injected into mice. Phal'lllacol Res Commun 1979; II: 759-73.

9. Toffano G. Benvegnu D. Bonetti AC. et al. Interaction of GMI ganglioside with crude rat brain neuronal mem­branes. J Neurochem 1980;35:861-6.

10. Carolei A. Fieschi C. Bruno R. Toffano G. Monosialogan­glioside GMI in cerebral ischemia. Cerebrovasc Brain Metab Rev 1991;3:134-57.

II. Tan WKM. Williams CE. Gunn AJ. Mallard EC, Gluckman PD. Pretreatment with monosialoganglioside GMI pro­tects the brain of fetal sheep against hypoxic-ischemic injury without causing systemic compromise. Pediatr Res 1993;34: 18-22.

12. Hayashi M. Satoh J. Sakamoto K. Morimatsu Y. Clinical and neuropathological findings in severe athetoid cerebral palsy: a comparative study of globo-Luysian and thalamo­putaminal groups. Brain Dev 1991;13:47-51.

13. Williams CEo Gunn AJ. Gluckman PD. Time course of intracellular edema and epileptiform activity following prenatal cerebral ischemia in sheep. Stroke 1991;22:516-21.

14. Mallard EC. Gunn AJ. Williams CE. Johnston BM. Gluck­man PD. Transient umbilical cord occlusion causes hip­pocampal damage in the fetal sheep. AMJ OIlS.El GYNECOL 1992; 167: 1423-30.

15. Siesjo BK. Bengtsson F. Grampp W. Theander S. Calcium. excitotoxins. and neuronal death in the brain. Ann N Y Acad Sci 1989;568:234-51.

16. de Boer J. Klein HC. Postema F. Go KG. KorfJ. Rat striatal cation shifts reflecting hypoxic-ischemic damage can be predicted by on-line impedance measurements. Stroke 1989;20: 1377-82.

17. Mahadik SP. Bharucha VA, Stadlin A. Oniz A, Karpiak SE. Loss and recovery of activities of a + and a isozymes of (Na' + K' )-ATPase in cortical focal ischemia: GMI gan­glioside protects plasma membrane structure and func­tion. J Neurosci Res 1992;32:209-20.

18. Skaper SD. Leon A, Facci L. Death of cultured hippocam­pal pyramidal neurons induced by pathological activation of N-methyl-o-aspartate receptors is reduced by monosia­logangliosides. J Phallllacol Exp Ther 1991 ;259:452-7.

19. Lipaniti M. Lazzaro A. Zanoni R. Mazzari S. Toffano G. Leon A. Monosialoganglioside GMI reduces NMDA neu­rotoxicity in neonatal rat brain. Exp Neurol 1991;113: 301-5.

20. Vaccarino F. Guidotti A. Costa E. Ganglioside inhibition of glutamate-mediated protein kinase C translocation in pri­mary cultures of cerebellar neurons. Proc Nat! Acad Sci USA 1987;84:8707-11.

21. Cuello AC. Garofalo L. Kenigsberg RL. Maysinger D. Gangliosides potentiate in vivo and in vitro effects of nerve growth factor on central cholinergic neurons. Proc Nat! Acad Sci USA 1989;86:2056-60.

22. Gluckman PD. Williams CE. Is the cure worse than the disease? Caveats in the move from the laboratory to clinic. Dev Med Child Neurol 1992;34: 1015-8.

23. Shen Y. Isaacson RL. Smothe. man WP. The behavioral and anatomical effects of prenatal umbilical cord damping in the rat and their alteration by the prior administration of nimodipine. Restor Neurol Neurosci 1991 ;3: 11-22.

24. Jensen A, KIossek H. Effects of flunariline on fetal circu­latory centralisation during acute asphyxia in sheep. In: Proceedings of the Society for the Study of Fetal Physiol­ogy. Asimolar. USA. 1990;17:0-10.

25. Levene MI. Gibson NA. Fenton AC. Papathoma E. Barnett D. The use of a calcium channel blocker. nicardipine. for severely asphyxiated newborn infants. Dev Med Child Neurol 1990;32:567-74.