progress review: hypoglycemic brain damag · progress review: hypoglycemic brain damag699 e roland...

Progress Review: Hypoglycemic Brain Damage 699

ROLAND N. AUER, M.D. , P H . D .

SUMMARY The central question to be addressed in this review can be stated as "How does hypoglycemiakill neurons?" Initial research on hypoglycemk brain damage hi the 1930s was aimed at demonstrating theexistence of any brain damage whatsoever due to insulin. Recent results indicate that uncomplicatedhypoglycemia is capable of killing neurons in the brain. However, the mechanism does not appear to besimply glucose starvation of the neuron resulting in neuronal breakdown. Rather than such an "Internalcatabotk death" current evidence suggests that in hypoglycemia, neurons are killed from without, i.e. fromthe extracellular space. Around the tune the EEG becomes isoelectric, an endogenous neurotoxm is pro-duced, and is released by the brain into tissue and cerebrospinal fluid. The distribution of necrotk neuronsis unlike that hi ischemia, being related to white matter and cerebrospinal fluid pathways. The toxin acts byfirst disrupting dendritic trees, sparing intermediate axons, indicating it to be an excltotoxln. Exactmechanisms of excitotoxk neuronal necrosis are not yet clear, but neuronal death involves hyperexcitation,and culminates in cell membrane rupture. Endogenous exdtotoxlns produced during hypoglycemia mayexplain the tendency toward seizure activity often seen clinically. The recent research results on which thesefindings are based are reviewed, and clinical implications are discussed.

Stroke Vol 17, No 4, 1986

THE DISCOVERY OF INSULIN in 1921 generatedthe initial interest in the possibility of brain damageresulting from hypoglycemia. This was due firstly tothe occurrence of insulin overdosage in the treatmentof diabetes mellitus. The second reason for this interestwas the deliberate administration of insulin in doseshigh enough to produce coma ("insulin shock" or insu-lin coma"), for amelioration of schizophrenia and oth-er psychiatric conditions. This treatment modality wasintroduced by Sakel in 19331-2 and led many to suspectthat a side effect, or indeed, the mechanism of actionof such "therapy" was due to resulting brain damage.Insulin was also given for acne and for morphine ad-diction.13

The pressing clinical issue of the appropriateness ofinsulin administration for psychiatric disease generat-ed heated controversy. The stage was thus set for aperiod of intense experimentation in the 1930s. Thesephysiologically uncontrolled experiments were de-signed to address the question of whether "permanentbrain damage" could result from insulin adminis-tration.

Early authors tacitly defined "permanent brain dam-age" as neuronal loss and gliosis. Such changes wereeasily demonstrable after the long recovery periodsused in these early experiments, ranging from days toweeks before insulin administration and sacrifice. Thevalidity of such a definition of brain damage seemsself-evident, since neuronal loss and gliosis constitutebrain damage by any pathologic criteria. Yet the darkand light neurons which have been recently demon-strated in hypoglycemia followed by no recovery orshort recovery periods4'' have aroused controversy6'7

as they do not meet such generally accepted pathologiccriteria for "brain damage".

The rejection of dark and light neurons as spuriousobservations was due to the well known occurrence ofsuch cells in poorly fixed central nervous tissue, which

From the Department of Pathology and Clinical Neurosciences, Uni-versity of Calgary, Calgary, Alberta, Canada.

Address correspondence to: Dr. Roland N. Auer, Health SciencesCenter, 3330 Hospital Drive N.W., Calgary, Alberta, Canada T2N 4N1

Received November 22, 1985; revison #1 accepted April 1, 1986.

led to the unquestioned assumption that all dark andlight neurons in experimental nervous tissue are due todelayed or inadequate fixation.6-8 It now appears thatthis assumption is not valid. Dark neurons are seen inthe brain acutely after excitotoxin administration9'l0 orafter ouabain injection,"-12 neither of which hamperscerebral perfusion. Excitotoxins and dark neurons willbe discussed more thoroughly below, but at present itwill suffice to emphasize that dark neurons representearly and reversible neuronal injury13 of a nonspecificnature.

Early workers thus used recovery periods longenough to show neuronal fallout and gliosis, ratherthan transient and early neuronal alterations. Howev-er, they were unable to identify causative factors in theinitial pathogenesis of neuronal necrosis. Remarkably,blood glucose levels were not reported in any of theseearly experiments, nor were physiologic parameters.Insulin itself was felt to be the substance harming thebrain, epitomized by one set of experiments in whichinsulin was injected into the brain, both ante and postmortem.14 In spite of such shortcomings in experimen-tal design, neuronal necrosis was clearly demonstratedin these early experiments, and its distribution wascarefully described.

In the cerebral cortex, most workers described asuperficial predilection for damage,13'l6 although nointerpretations of this were made. If hypoglycemiacauses superficial cortical damage, with preservationof the remaining cortex17' " clinical cases of irrevers-ible post-hypoglycemic coma which demonstrate con-tinued cerebral blood flow19 becomes less surprising.Continued perfusion of intact deeper cortical layerswould be expected.

In the hippocampus, two research groups observed arelationship of necrotic neurons to CSF spaces.2021

The American neuropathologist Arthur Weil and hiscollaborators noted involvement of the dentate gymsnear the ventricle, and explicitly stated that in hypogly-cemic brain damage, toxic substances in the cerebro-spinal fluid should be hypothetically considered,21 aswell as the dominant vascular22 and neuronal23 theoriesof Spielmeyer and the Vogts, respectively. This pre-

by guest on May 2, 2017

http://stroke.ahajournals.org/D

ownloaded from

http://stroke.ahajournals.org/

700 STROKE VOL 17, No 4, JULY-AUGUST 1986

scient observation was then forgotten or overlooked bysubsequent workers.

Since the results of this early period of intensiveresearch had clearly demonstrated neuronal necrosis,and since the response rate of schizophrenia to insulincoma therapy was disappointing,24 insulin shock treat-ment gradually fell out of common use. Research inter-est in the structural aspects of hypoglycemic braindamage waned thereafter. Hypoglycemia was consid-ered to be a form of ischemia,23"27 and the two insultswere described as having the same neuropathol-OgV 27.2« y ^ tacit assumption was that in conditions ofeither oxygen or glucose deprivation, internal defi-ciencies of neuronal energy metabolism developed,resulting in selective necrosis of neurons before glia.Neurons were felt to be selectively vulnerable becauseof their very active metabolism. This view prevaileduntil accumulating neurochemical data prompted areexamination of the tenet that ischemic and hypogly-cemic brain damage are identical deficiency states ofthe brain.

Neurochemical ResultsIt soon became apparent that there were consider-

able problems associated with the concept that ische-mia and hypoglycemia were identical insults, produc-ing neuronal necrosis by either substrate (glucose) orelectron acceptor (oxygen) deficiency.

Neurochemical whole-brain analyses revealed pro-found differences between the two insults: in manyparameters, the deviations from normal brain wereactually in opposite directions. For example, cellularredox systems are reduced in ischemia,29 but are oxi-dised in hypoglycemia.30 Although brain pH is de-creased in ischemia due to the formation of lactic acid,it is increased in hypoglycemia.3132 This alkalosis isdue both to the formation of ammonia from deamina-tion of amino acids,433"3* and to the consumption ofmetabolic acids and the absence of lactic acid forma-tion. Although alkalosis has been held to be directlyresponsible for acidophilic neuronal death,37 reportedvalues for intracellidar pH in hypoglycemia have dif-fered among research groups. Derived values calculat-ed from intracellular equilibria of weak acids suggestan alkalosis,3132 conflicting with recent NMR data.38

Energy failure, ion fluxes, and lipolysis are com-mon to both ischemia and hypoglycemia.39 However,energy failure is considerably less severe in hypogly-cemia than in ischemia, in spite of the normal depend-ence of the brain on both oxygen and glucose, whichare consumed in a stoichiometric relationship.39 De-pression of cerebral oxygen consumption in hypogly-cemia is less than expected in proportion to depressionof glucose utilization, indicating the considerablemetabolic versatility of the brain in oxidizing non-glucose fuels to maintain its energy state.30' 4O~46

Even lactic acid can be burned as fuel,47 reversinghypoglycemic stupor,48 but this functions only in im-mature animals due to the failure of blood lactate toenter the CNS in mature animals.49 However, produc-tion of lactic acid during hypoglycemia is not possible

due to glucose deficiency. This likely explains whyinfarction is not a feature of uncomplicated hypoglyce-mia (see below).

In hypoglycemia profound enough to cause cessa-tion of electrical activity ("isoelectricity"), ATP levelsare still over one third of the normal val-ues34,36.45,46,50-53 ̂ ue t o oxidation of endogenous sub-strates (proteins and fats) for fuel30'46'53"55, whereas inischemia ATP drops to less than 5% of normal.3956

Lowering of the blood pressure during hypoglycemiaresults in no increase in neuronal necrosis,37 in spite ofincreased cellular release of potassium,58 and enhancedenergy failure.57 Together, these results indicate thatcerebral energy failure per se, as measured in wholebrain analyses, does not account for the phenomenonof selective neuronal necrosis within the brain. Selec-tive necrosis of neurons, sparing glia, seems instead tobe related to the presence of excitatory receptors (seebelow) on neurons but not glia. Astrocytes and oligo-dendrocytes are also exposed to hypoglycemia andenergy failure, and in spite of being metabolically ac-tive cells, demonstrate reactive changes but not necro-sis in hypoglycemia.59

When hypotension complicates hypoglycemia, allareas of the brain are not exposed to an equal fall inperfusion. Due to focal loss of autoregulation inducedby the hypoglycemia, some regions are grossly under-perfused whereas others show almost no fall in localcerebral blood flow.60 Glucose delivery to the mosthypoperfused areas should be most compromised.However, tissue damage is not at all exacerbated inthese regions. For example, the cingulate gyms, se-verely hypoperfused when hypotension complicateshypoglycemia,60 does not show focal intensification inthe density of neuronal necrosis.37 Blood flow is evi-dently not a critical determinant of hypoglycemia-in-duced neuronal necrosis.

The ionic and chemical changes in hypoglycemiatake place abruptly at the onset of cerebral EEG isoe-lectricity.*-34-36'45'46'51'53'55'60-64 A prominent changein cerebral amino acids seen in hypoglycemia is a 3 to4-fold rise in tissue aspartate and a fall in gluta-m a t e / , 34,38,65-68 jjyg metabolic effect is related to ei-ther protein breakdown69"72 and/or lack of protein syn-thesis,73' 74 with pooling of amino acids and a shift inthe aspartate-glutamate transaminase reaction towardaspartate. The rise in tissue aspartate is of importancesince this compound is a known neurotoxin by virtueof its excitatory properties ("excitotoxin", see below).

Metabolically derived aspartic acid is also markedlyincreased when "focal hypoglycemia" is produced inthe brain by local administration of haloacetate com-pounds, which poison glycolysis,66' "•76 This increasein aspartate occurs in spite of normal serum glucoselevels, indicating the metabolic abnormality to be dueto inhibition of glycolysis, rather than to low glucoselevels per se.

In view of the summarized data that the onset ofionic and metabolic changes of hypoglycemia occurabruptly at the onset of cerebral EEG isoelectricity, itwill come as no surprise that neuronal necrosis is ab-



ownloaded from


HYPOGLYCEMIC BRAIN DAMAGEMu«r 701

sent in hypoglycemia unless the EEG becomes isoelec-tric, regardless of the blood sugar level at which theEEG goes flat.17 Furthermore, the duration of electro-cerebral silence roughly determines the degree of re-sultant brain damage.17

Blood-to-brain glucose transfer is effected chieflyby carrier mediated facilitated transport.77"79 However,the metabolic changes outlined above occur to a muchlesser extent in the cerebellum than in the cere-brum. 62-73-74' *°-n This may be due to a more efficientblood-to-brain glucose transport system81 or a densercapillary network79 in the cerebellum, giving rise to alesser metabolic perturbation. Structural damage in thecerebrum far exeeds that in the cerebellum,18 in paral-lel with the metabolic findings.73748082

Recent Morphological ResultsStructural analysis in comparable rat models with

long term survival have distinguished hypoglycemia asa brain insult18 unique from ischemia83 or epilepsy.84

Indefinite survival is possible even after 60 minutes ofelectrocerebral silence due to hypoglycemia in therat.17 Thirty minutes of hypoglycemic isoelectricitygives rise to few necrotic neurons.17 These results inhypoglycemia are in accord with clinical experience:

. »ss-

FIGURE 2. Acidophilic neuronal necrosis of dentate granuleneurons at the crest of the dentate gyms (arrows), facing thecistern velum interposition. Neurons with dendritic trees ex-posed to the cistern are necrotic (black and punctate in thephotomicrograph). Remaining granule cells of the internal andexternal blades lie with their dendrites under the fused hippo-campal fissure, or opposed to the diencephalon, respectively.With a greater hypoglycemic insult, these cells become affectedas well. Rat, 30 min hypoglycemia + 1 week recovery. AcidfuchsinlCresyl violet. Bar = 500 fun.

\

p*

23

5

6

FIGURE 1. Cerebral cortex 1 week after 60 min with a flatEEG due to hypoglycemia. Necrotic neurons, marked by blackpunctate spots (arrow) in this photomicrograph, predominate inthe upper cortical layers, chiefly layers 2 and 3. Layer 4 isalmost normal, and the large pyramidal neurons of layer 5 arewell preserved. Acid fuchsinlCresyl violet. Bar = 500 fim.

The desired period of "insulin coma" during the periodwhen this therapy was in vogue was 30 to 60 min.

The distribution of brain damage is also unique inhypoglycemia. Necrotic neurons in several brain re-gions show a conspicuous relationship in the rat to thecisterns of the subarachnoid space.18 In the rat cerebralcortex (fig. 1), this relationship is seen as a superficialto deep gradient in necrotic neurons: layer 2 is moreaffected than layer 3, which is in turn more affectedthan layer 4. It should be noted that neurons in affectedcortical laminae include both pyramidal and granulecell types. Thus, cellular necrosis in the cerebral cortexin hypoglycemia is not limited to one type of neuron.

In the hippocampus proper, neuronal necrosis isdenser medially, in a gradient along the CA1 pyrami-dal neurons (Sommer's sector) and subiculum, whichform a continuous cell band. Cells die within twohours medially where damage is greater than lateral-ly.59 A lingering neuronal death often occurs lateral-ly.59 With severe insults, all neuronal types are de-stroyed, including the "resistant" CA3 cells.59

The dentate gyrus of the hippocampus shows consis-tent necrosis in hypoglycemia.83 As in the pyramidalcell band of the hippocampus, however, all dentateneurons are not equally affected. Granule cells relatedto the subarachnoid cisterns, for example those at thecrest of the gyrus (fig. 2), facing the cistern veluminterpositum, show early and conspicuous necrosis.The location of the dendritic tree, and not merely thecell body, appears to be a critical factor. Cell bodieswith their dendrites facing the cisterna ambiens under-go necrosis whereas cells with dendrites under thefused hippocampal fissure are relatively spared.85

Studies of the time course of damage in various



ownloaded from



brain regions show that the cortex and hippocampusshow early damage already after 10 rain cerebral iso-electricity,1359 whereas the caudate nucleus is normalfor one hour following isoelectricity.86 Dark neurons,and then acidophilic neurons finally appear in the cau-date, but develop later than in the cortex. Together,these facts suggest that a neurotoxin is carried into thecaudate from the cerebral cortex, via the white materbundles, giving rise to dark, and then necrotic neurons.

Results in other brain regions are in accord withthose in the cerebral cortex, hippocampus, and caudatenucleus: Homogeneous neuronal populations are notuniformly affected. Rather, damage is located in welldefined locations often related to CSF spaces and whitematter. For example, of the nuclei in the amygdaloidcomplex, only those portions near the lateral ventricleshow necrosis, and in the cerebellum, Purkinje cellnecrosis is seen in the folia facing the foramena ofLuschka.18

With such evidence implicating a toxic substance inand around the brain, the question naturally arises as tothe possible nature of such a toxin. The followingboundary conditions are imposed by the neuropath-ology:1. The substance must produce selective neuronal ne-

crosis. This is to say it should be toxic only toneurons and not to glia or other cell types, sinceselective neuronal necrosis but not infarction isseen in pure, controlled hypoglycemic insults. Thesubstance is thus a neurotoxin, not a histotoxin.

2. The neurotoxin must be capable of killing all neu-ronal types in the hippocampus, cerebral cortex,caudate nucleus and spinal cord, as necrosis of allthese neuronal elements has been observed aftercontrolled experiments delivering a "pure" hypo-glycemia insult. Dentate granule cells and CA1pyramidal cells must be more sensitive than CA3pyramidal cells in the hippocampus. Small and me-dium sized neurons should be more vulnerable thanlarge neurons in the caudate.

Molecules fulfilling these two basic criteria exist,and are termed excitotoxins, due to their parallel exci-tatory and neurotoxic properties." They bind to recep-tors located on neuronal dendritic trees and perikarya,causing increased neuronal spiking and neuronal de-polarization.87 The sequence of critical events subse-quently leading to neuronal death are as yet unclear,but there is no question that these compounds are capa-ble of killing neurons. The characteristic neuropathol-ogy is destruction of dendrites in the neuropil due tothe location of receptors for these substances on thedendrites, while axons, which lack such receptors, arespared. Selective neuronal necrosis results when cellmembrane destruction spreads from dendrites to neu-ronal perikarya. The glia survive.

Is there evidence in hypoglycemia that the neuro-toxin is an excitotoxin? Recent results give an affirma-tive answer.83- 88^° Examination of the dendritic treesof hippocampal neurons destined to undergo necrosisreveals early dendritic, axon-sparing lesions (fig. 3).Swollen dendrites, containing swollen mitochondria

-> &,

FIGURE 3. Axon-sparing, dendritic lesion. Segmental dilata-tion of a dendrile is seen, but the surrounding neuropil is intact.Synapses (s) indicate the dendritic nature of the swollen pro-cesses. Swollen mitochondria showing fractured cristae arecommon within such dendritic processes (inset). Hippocampus,subiculum, rat, 10 min EEG isoelectricity induced by hypogly-cemia. Bar = / fim.

(fig. 3, inset), appear after 10 minutes of cerebral EEGisoelectricity. Neuronal cell bodies are normal at thistime.

By 30 minutes, swollen mitochondria and dark celltransformation are seen in the cell body, where cellmembrane irregularities appear. One hour after theonset of electrocerebral silence, the neurons arenecrotic, indicated by mitochondrial flocculent densi-ties and frank membrane rupture. The cytoplasm ofthese cells stains acidophilic by light microscopy, andthere is free admixture of amorphous cytoplasm withthe extracellular space (fig. 4), indicating cellular dis-solution (cytolysis).

The early sequence of cellular changes thus indi-cates a "dendritic death" of neurons characteristic ofexcitotoxins. The entire sequence of changes is sum-marized in figure 5. The characteristic ultrastructuralappearance is that of axon-sparing, dendro-somal pa-thology.

¥&•

FIGURE 4. Cytorrhexis, occurring after four hours, indicatedby rupture of cell membranes and admixture of amorphouscytoplasmic contents with the extracellular space (left). Theremnant of the nucleus and a few surrounding mitochondria,are on the right. Bar = 5 fim.



ownloaded from


HYPOGLYCEMIC BRAIN DAMAGEJAuer 703

Removal of the excitatory cortico-striatal inputs tothe caudate by cortical ablation, a procedure whichwould lead to degeneration of excitatory amino acidreceptors on caudate neurons, virtually abolishes neu-ronal necrosis in the caudate nucleus.89 Furthermore,removal of the dopaminergic input to the striatum with6-OH dopamine lesions of the substantia nigra alsoreduces selective neuronal necrosis in the caudate.90

The latter procedure does not even remove a pathwayusing an excitotoxic amino acid (eg, glutamate, aspar-tate) as a neurotransmitter, but merely one which maymodulate neuronal excitation in the caudate (See Lind-vall et al for discussion.)90 Clearly, the degree of exci-totoxic neuronal necrosis is affected by inhibitory andexcitatory inputs.

Pharmacologic blockade of excitotoxin receptorsalso prevents neuronal necrosis.88'91 Together with themorphologic findings, these results lend considerablesupport to an excitotoxic mechanism of neuronal deathin hypoglycemia, rather than catabolism due to a glu-cose deficiency state directly killing the neuron.

ConclusionsIt is possible to draw eight conclusions regarding

hypoglycemic brain damage, which can now be sum-marized:1. In the rat, neuronal necrosis is absent in hypoglyce-

mia unless the EEG becomes isoelectric.2. Quantitated neuronal necrosis increases as the du-

ration of EEG isoelectricity increases.3. Both the density and the distribution of brain dam-

age is unique in hypoglycemia.4. Although different neuroanatomical cell types are

differentially susceptible to necrosis, eventually allneuronal types can undergo necrosis with a highdose of insult.

5. The same neuroanatomical cell type may be onlypartially affected in a "subanatomic distribution".

6. In the rat, with progressively longer periods of isoe-lectricity, neuronal necrosis appears first in sus-ceptible brain regions exposed to subarachnoidspaces, second in portions of brain apposed to otherbrain surfaces, and lastly in deeper regions. Thissuggests a toxic substance accumulating in theCSF.

7. The fact that only selective neuronal necrosis andnot infarction is seen in even severely damagedbrains after purely hypoglycemic insults indicatesthe hypoglycemic toxin is a neurotoxin, and not ahistotoxin.

FIGURE 5. Sequence of events in "dendriticdeath" of neurons caused by excitotoxins, asdemonstrated in hypoglycemia. Swelling ofdendrites begins the process, and cell mem-brane dissolution in the dendritic tree andperi-karyon signal cell death.

8. The axon-sparing, dendrosomal nature of the neu-ronal lesion, and its prevention by antagonists ofexcitotoxin receptors, indicates that the hypoglyce-mic neurotoxin is an excitotoxin.

SynthesisA synthesis of the likely events leading to neuronal

death in hypoglycemia can now be made. As the bloodglucose falls to the range of 2.0-2.5 mM, slowing ofthe EEG appears. As energy failure progresses, ionicpumps cannot be maintained and ions run down theirconcentration gradients. These events begin focally,but spreading depression of cortical activity occurs,often asynchronous within or between the hemi-spheres. EEG silence ensues. A new metabolic ho-meostasis is achieved, with an ATP level of roughly25% of control, maintained due to oxidation of endog-enous substrates within the brain. The aspartate-gluta-mate reaction reaches a new steady state, the equilibri-um being markedly shifted toward aspartate. Thismetabolically-derived aspartate is released into the ex-tracellular space of the brain, first into the interstitialspace and ultimately into the CSF.

The question naturally arises as to why there shouldbe a relationship of necrotic neurons or their dendritesto CSF spaces, if an excitotoxin is endogenously pro-duced by the brain parenchyma. It is known that theinterstitial fluid of the brain is derived from the arterialend of the microcirculation at the capillary-glia com-plex,92 most of it undergoing resorption into the bloodstream at the venular end of the microcirculation. Thebalance of unresorbed interstitial fluid courses throughthe brain parenchyma93 toward the CSF to be ultimate-ly resorbed by the arachnoid villi. Dilution of anyexcitotoxin produced and locally released by the braininto the CSF occurs only via mixing with CSF from thedistant choroid plexus. Cerebral interstitial fluid, incontrast to CSF, is renewed in closer proximity to itssource at the arterial end of the microcirculation byblood pressure driven production.94 Thus, an endogen-ously produced excitotoxin might still appear in higherconcentration in the CSF than at its site of production,due to proximal "washout" and constant renewal ofcerebral interstitial fluid. Subarachnoid CSF, especial-ly that surrounding the spinal cord, is known to berelatively stagnant due to relative remoteness fromchoroid plexus tissue, and poor exchange of spinal andcerebral CSF.

The released excitotoxin, likely aspartate, then



ownloaded from



binds to neuronal dendrites, initiating a lethal chain ofcellular events. Further experiments are necessary toclarify the exact events responsible for excitotoxicneuronal death, and their sequence.

Future ResearchHow do excitotoxins kill neurons? Once an excito-

toxin binds to neuronal dendrites, definite mechanismsleading to neuronal death are still unclear, but ionicfluxes probably play a crucial role. Excitotoxins causean influx of Ca+ + ions into neurons,93'** with calciumdependent neurotoxic effects.97 Lipolysis is cactivatedby calcium.39-M Should degradation of dendritic cellmembranes occur in this process of Ca+ + activatedlipolysis, then neurons would be effectively open tothe extracellular space and unable to compartmentalizecytoplasm. Such a sequence may occur in excitotoxicneuronal death.

Excitotoxins can be classified into those which pro-duce only local lesions, and those which produce bothlocal and "distant" lesions,9 the latter arising in a brainregion remote from an injection site. The distant le-sions appear to arise by transsynaptic mechanisms, andare seen with excitotoxins capable of producing a sei-zure-related brain damage (SRBD) syndrome. The lat-ter consists of severe status epilepticus and selectiveneuronal necrosis in a trans-synaptic distribution. Incontast to findings in epilepsy, where a trans-synapticdistribution is seen,84 no brain regions show a trans-synaptic distribution in hypoglycemia. An SRBD syn-drome has not been seen in experimental studies ofhypoglycemia.1720"-102 An excitotoxin with a pre-dominantly local effect is thus implicated.

Increased aspartic acid levels in brain tissue duringhypoglycemia,4'*• 65~68'16 and a largely local neuronaldestruction resulting from injections of aspartate intobrain tissue,103 make endogenously produced asparticacid the most likely excitotoxic agent mediating theneuronal necrosis in hypoglycemia. However, quino-lini • acid, an endogenous cerebral tryptophan metabo-lite, :s lethally excitotoxic in nanomole quantites104- m

and might also be produced during hypoglycemia.Quinolinic acid levels have not yet been measuredduring hypoglycemia.

In hypoglycemia, it is necessary to measure cerebralinterstitial fluid levels of excitotoxins using either in-tracerebral microdialysis or push-pull cannulae, com-paring them with CSF levels of excitotoxins. The obvi-ous experiment suggested by the data available istransfer of either cerebral interstitial fluid or cerebro-spinal fluid from a hypoglycemic rat into the brain of arecipient animal. If excitotoxic pathology could bedemonstrated by light microscopy (selective neuronalnecrosis), and by elecron microscopy (ie an axon-spar-ing lesion), this would constitute definite evidence fora fluid-borne excitotoxin in such a "transfer bioassay"experiment.

Neonatal HypoglycemiaNeonatal hypoglycemia in all probability constitutes

an entirely different situation from the adult with re-

gard to hypoglyemic brain damage. Metabolically, theimmature brain has the capability to oxidize lactate forfuel,48'l06 an apparent protective response to neonatalhypoglycemia.l07

Pathologic responses in the immature brain are alsodifferent from that seen in adult animals. The dark celltransformation in aldehyde fixed tissue, a non-specificneuronal reaction to excitotoxin administration,9'l0

ouabain administration,11'l2 or mechanical injury priorto fixation108 is not seen prior to a certain stage ofneuronal maturation.109 Thus, the dark neuron, a cellu-lar response of the brain seen acutely in hypoglyce-mia,4' '•l3 and likely representing neuronal ion and wa-ter shifts, does not occur in the neonatal brain.

Excitotoxic neuronal death would be theoreticallyimpossible prior to developmental maturity of neuro-transmitter systems and cell surface receptors. Hypo-glycemia to less than 20 mg/100 ml usually gives riseto cerebral EEG isoelectricity and consequently braindamage in the mature brain.17 In the neonate, glucoselevels in this range uncommonly give rise to a perma-nent neurologic deficit if the hypoglycemic episode isasymptomatic.110 Work is needed specifically on neo-natal hypoglycemia and at present, an excitotoxicmechanism should not be assumed by transfer of find-ings obtained in adult animals to neonates. The fewstudies that have addressed the problem of neonatalhypoglycemia have demonstrated structural braindamage in both man1"- "2 and experimental animals,113

the mechanism of which remains unclear.

Clinical and Medico-Legal AspectsAlthough a SRBD syndrome of continuing and in-

tractable seizures does not characterize hypoglycemia,it is well known clinically that seizures are a manifesta-tion of profound hypoglycemia, prior to the stage ofcoma. When the EEG becomes isoelectric, corre-sponding to clinical coma, the energy failure and corti-cal depolarization make seizure manifestations impos-sible. But the onset of EEG isoelectricity is not asimultaneous event over the cerebral hemisphere.Rather, the wave of ionic shifts and cortical depolar-ization spread over the hemisphere,53'63 not unlike cor-tical spreading depression. Portions of the cortexwhich are already isoelectric may well release excito-toxin before an adjacent cortical region. The resultingexcitation, possibly combined with the depletion of theinhibitory neurotransmitter GABA,34"36'll4' "5 may ac-count for seizure phenomena in hypoglycemia.

Focality of initial electrical events and energy fail-ure may also explain the focal symptoms often seen inhypoglycemia, when no focal perfusion deficit canoffer a satisfactory explanation."6 Another explana-tion for focal symptoms in hypoglycemia may be focalhypoperfusion,117 probably operating via decreasedglucose delivery to the tissue. However, it is to beemphasized that focal hypoperfusion in hypoglycemiacan account for reversible functional deficits, "6 but notneuronal necrosis.37

Although amyotrophy in diabetic patients may bedue to peripheral neuropathy of the axonal type, rather



ownloaded from


HYPOGLYCEMIC BRAIN DAMAGE/Auer 705

than to anterior horn cell damage, necrosis of the mo-tor neurons of the spinal cord has been reported afterprofound hypoglycemia.17'18' I I7I2° Lower motor neu-ron paresis develops slowly and is delayed after theepisode of hypoglycemia,"7' '"•121 possibly reflectinga delayed excitotoxic death of susceptible motor neu-rons of the anterior horn1 7 'm due to accumulation of anexcitotoxic amino acid in the stagnant CSF surround-ing the spinal cord.

It is apparent that structural damage to the nervoussystem in humans resembles that seen in animals, ap-pears independent of the agent causing the hypoglyce-mia, having been seen in patients after oral hypoglyce-mic agents123-l24 or islet cell adenoma of thepancreas,3-117> "9 in addition to insulin overdose in dia-betes.3' 123~127 The etiology of the hypoglycemia is thusnot important in the immediate clinical situation, andclinical efforts are best directed toward rapid return ofplasma glucose levels to normal and avoidance of cere-bral EEG isoelectricity.17 Death due to severe braindamage from therapeutic "insulin shock" for psychiat-ric disease128"130 is fortunately no longer seen.

The third conclusion listed above may have medico-legal implications: If a superficial laminar necrosis ofneurons is demonstrable in the cerebral cortex in anautopsy case of possible insulin overdosage, as hasbeen demonstrated123-124-13° then hypoglycemic braindamage is very likely, as this pattern has not beenreported after ischemia. A laminar distribution ofneuronal necrosis to the middle cortical laminae wouldfavor ischemic brain damage, and a transcortical dis-tribution of necrotic neurons would be unhelpful indistinguishing ischemic from hypoglycemic braindamage.

The absence of infarction in experimental models ofhypoglycemic brain damage may tempt one into attrib-uting infarction only to primary ischemic insults.However, when glucose is administered followingprofound hypoglycemia, there is considerable post-hypoglycemic hypoperfusion.131 More importantly,lactic acid levels rise to supra-normal levels34'M'M inthe tissue. As high lactic acid levels in brain tissueappear to promote infarction,13213* it may be judiciousto administer glucose slowly and continuously whenrecovering clinical cases of hypoglycemic coma, inorder to prevent infarction in the recovery phase. TheEEG is useful in monitoring recovery, returning inminutes after a mild hypoglycemic insult, but afterlonger periods following a more prolonged period ofhypoglycemia.17'137

The presence of absence or infarction should thusnot be used by the neuropathologist in a medico-legalsituation in attempting to differentiate, with certainty,ischemic from hypoglycemic brain damage. Infarctionmay well develop secondarily in the recovery periodafter hypoglycemia, due to tissue lactic acidosis aftertoo rapid glucose administration.

AcknowledgmentsThe author is grateful to Drs. Bo K. SiesjS, Tim Watson and Rolando

Del Maestro for comments on the manuscript, and to Caroline Collinsfor typing the manuscript.

References1. Sakel M: Neue Behandlungsart Schizophreniker und verwirrter

Erregter. Wien klin Wchnschr 46: 1372-1373, 19332. Sakel M: The methodical use of hypoglycemia in the treatment of

psychoses. Amer J Psychiat 94: 111-129, 19373. Baker AB: Cerebral lesions in hypoglycemia. II. Some possibili-

ties of irrevocable damage from insulin shock. Arch Path 26:765-776, 1938

4. Agardh C-D, Kalimo H, Olsson Y, Siesjo BK: Hypoglycemicbrain injury. I. Metabolic and light microscopic findings in ratcerebral cortex during profound insulin-induced hypoglycemiaand in the recovery period following glucose administration. ActaNeuropathol (Beri) 50: 31—41, 1980

5. Kalimo H, Agardh C-D, Olsson Y, Siesjd BK: Hypoglycemicbrain injury. II. Electron microscopic findings in rat cerebralneurons during profound insulin-induced hypoglycemia and in therecovery period following glucose administration. Acta Neuro-pathol (Beri) 50: 43-52, 1980

6. Brierley JB, Brown AW: Remarks on the papers by C-D Agardhet al/H. Kalimo et al. "Hypoglycemic brain injury, I, II". ActaNeuropathol (Beri) 55: 319-322, 1981

7. Agardh C-D, Kalimo H, Olsson Y, Siesjo BK: Reply to theremarks by JB Brierley and AW Brown. Acta Neuropathol (Beri)55: 353-325, 1981

8. Cammermeyer J: The importance of avoiding "dark" neurons inexperimental neuropathology. Acta Neuropathol (Beri) 1:245-270, 1961

9. Olney JW: Excitotoxins: An Overview. In: Fuxe K, Roberts P,Schwarcz R (eds). Excitotoxins, Wenner-Gren InternationalSymposium Series, Vol 39. MacMillan Press Ltd, London, pp82-96, 1983

10. Honchar MP, Olney JW, Sherman WR: Systemic cholinergicagents induce seizures and brain damage in lithium treated rats.Science 220: 323-325, 1983

11. Cornog JL Jr, Gonatas NK, Feierman JR: Effects of intracerebralinjection of ouabain on the fine structure of rat cerebral cortex.Amer J Patbol 51: 573-590, 1967

12. Towfighi J, Gonatas NK: Effect of intracerebral injection of oua-bain in adult and developing rats. An ultrastructural and autora-diographic study. Lab Invest 28: 170-180, 1973

13. Auer RN, Kalimo H, Olsson Y, Siesjo BK: The temporal evolu-tion of hypoglycemic brain damage. I. Light and electron micro-scopic findings in the rat cerebral cortex. Acta Neuropathol (Berl)67: 13-24, 1985

14. Stief A, Tokay L: Weitere experimentelle Unteresuchungen fiberdie cerebrale Wirkung des Insulins. Z Ges Neural Psychiat 153:561-572, 1935

15. Grayzel DM: Changes in the central nervous system due to con-vulsions due to hyperinsulinism. Arch IntMed 54:694-701, 1934

16. Stief A, Tokay L: Beitrag zur Histopathologie der experimentel-len Insulinvergiftung. Z Ges Neural Psychiat 139:434-461, 1932

17. Auer RN, Olsson Y, Siesjd BK: Hypoglycemic brain injury in therat. Correlation of density of brain damage with the EEG isoelec-tric time: A quantitative study. Diabetes 33: 1090-1098, 1984

18. Auer RN, Wieloch T, Olsson Y, Siesj6 BK: The distribution ofhypoglycemic brain damage. Acta Neuropamol (Beri) 64:177-191, 1984

19. Agardh C-D, Rosin I, Ryding E: Persistent vegetative state withhigh cerebral blood flow following profound hypoglycemia. AnnNeural 14: 482-486, 1983

20. Finley KH, Brenner C: Histologic evidence of damage to the brainin monkeys treated with Metrazol and insulin. Arch Neural Psy-chiat (Chic) 45: 403-438, 1941

21. Weil A, Liebert E, Heilbrunn G: Histopathologie changes in thebrain in experimental hyperinsulinism. Arch Neural Psychiatr 39:467-481, 1938

22. Spielmeyer W: Pathogenese ortlich elektiver Gehimveranderun-gen. Z Ges Neurol Psychiatr 99: 756-776, 1925

23. Vogt C, Vogt O: Sitz und Wesen der Krankheiten im Lichte dertopistischen Hirnforschung und des Varierens der Tiere. J PsycholNeurol 47: 237-457, 1937

24. Mayer-Gross W: Insulin coma therapy of schizophrenia: Somecritical remarks on Dr. Sakel's report. J Ment Sci 97: 132-135,1951



ownloaded from


706 STROKE V O L 17, No 4, JULY-AUGUST 1986

25. CourvilleCB: Late cerebral changes incident to severe hypoglyce-raia (insulin shock). Their relation to cerebral anoxia. ArchNeurol Psychiatr 78: 1-14, 1957

26. Richardson JC, Chambers RA, Heywood PM: Encephalopathiesof anoxia and hypoglyceraia. Arch Neurol 1: 178-190, 1959

27. Brierley JB, Meldrura BS, Brown AW: The threshold and neuro-pathology of cerebral "anoxic-ischemic" cell change. ArchNeurol 29: 367-374, 1973

28. Brierley JB, Brown AW, Meldrum BS: The nature and timecourse of the neuronal alterations resulting from oligaemia andhypoglycemia in the brain of Macaca mulatta. Brain Res 25:483-499, 1971

29. Rehncrona S, Folbergrova J, Smith D, Siesjd BK: Influence ofcomplete and pronounced incomplete cerebral ischemia and sub-sequent recirculation on cortical concentrations of oxidized andreduced gluthathione in the rat. J Neurochem 34: 477-486, 1980

30. Norberg K, Siesjo BK: Oxidative metabolism of the cerebralcortex of the rat in severe insulin induced hypoglycemia. J Neuro-chem 26: 345-352, 1976

31. '"elligrino D, Siesjd BK: Regulation of extra- and intracellular pHL-. die brain in severe hypoglycemia. J Cereb Blood Flow Metabol1: 85-96, 1981

32. Pelligrino D, Almquist L-O, Siesjd BK: Effects of insulin-in-duced hypoglycemia on intracellular pH and impedance in thecerebral cortex of the rat. Brain Res 221: 129-147, 1981

33. KozlovNB: On the role of ammonia in the development of insulinshock. Vop Med Khim 6: 396-402, I960

34. Tews JK, Carter SH, Stone WE: Chemical changes in the brainduring insulin hypoglycemia and recovery. J Neurochem 12:679-683, 1965

35. Lewis LD, Ljunggren B, Norberg K, Siesjd BK: Changes incarbohydrate substrates, amino acids and ammonia in the brainduring insulin-induced hypoglycemia. J Neurochem 23:659-671, 1974

36. Agardh C-D, Folbergrova J, Siesjd BK: Cerebral metabolicchanges in profound insulin-induced hypoglycemia, and in therecovery period following glucose administration. J Neurochem31: 1135-1142, 1978

37. Vanderhaeghen JJ, Logan WJ: The effect of the pH on the in vitrodevelopment of Spielmeyer's ischemic neuronal changes. J Neur-opathol Exp Neurol 30: 99-104, 1971

38. Behar KL, den Hollander JA, Petroff OAC, Hetherington HP,Prichard JW, Shulman RG: Effect of hypoglycemic encephalop-athy upon amino acids, high-energy phosphates, and pH, in therat brain in vivo. Detection by sequential H and 3IP NMR spec-troscopy. J Neurochem 44: 1045-1055, 1985

39. Siesjd BK: Cell damage in the brain: A speculative synthesis. JCereb Blood Flow Metabol 1: 155-185, 1981

40. Kety SS, Woodford RB, Harmel MH, Freyhan FA, Appel KE,Schmidt CF: Cerebral blood flow and metabolism in schizophre-nia. The effect of barbiturate semi-narcosis, insulin coma andelectroshock. Amer J Psychiatr 104: 765-770, 1948

41. Fazekas JF, AI man RW, Parrish AE: Irreversible posthypoglyce-mic coma. Am J Med Sci 222: 640-643, 1951

42. Eisenberg S, Seltzer HS: The cerebral metabolic effects of acutelyinduced hypoglycemia in human subjects. Metabolism 11:1162-1168, 1962

43. Delia Porta P, Maiolo AT, Negri VU, Rossella E: Cerebral bloodflow and metabolism in therapeutic insulin coma. Metabolism 13:131-140, 1964

44. Pappenheimer JR, Setchell BP: Cerebral glucose transport andoxygen consumption in sheep and rabbits. J Physiol 233:529-551, 1973

45. Agardh C-D, Chapman AG, Nilsson B, Siesjo BK: Endogenoussubstrates utilized by rat brain in severe insulin-induced hypogly-cemia. J Neurochem 36: 490-500, 1981

46. Ghajar JBG, Plum F, Duffy TE: Cerebral oxidative metabolismand blood flow during acute hypoglycemia and recovery in unan-aesthetised rats. J Neurochem 38: 397-409, 1982

47. Mcllwain H: Glucose level, metabolism, and response to electri-cal impulses in cerebral tissues from man and laboratory animals.Biochem J 55: 618-624, 1953

48. Thurston JH, Hauhart RE, Schiro JA: Lactate reverses insulin-induced hypoglycemic stuor in suckling-weanling mice: Bio-

chemical correlates in blood, liver, and brain. J Cereb Blood FlowMetabol 3: 498-506, 1983

49. Cremer JE, Cunningham VJ, Pardridge WM, Braun LD, Olden-dorf WH: Kinetics of blood-brain barrier transport of pyruvate,lactate and glucose in suckling, weanling and adult rats. J Neuro-chem 33: 439-445, 2979

50. Olsen NS, Klein JR: Effect of insulin hypoglycemia on brainglucose, glycogen, lactate and phosphates. Arch Biochem 13:343-347, 1947

51. Hinzen DH, MOlIer U: Energiestoffwechsel und Funktion desKaninchengehims wahrend InsulinhypoglykSmie. Pflugers Archges Physiol 322: 47-59, 1971

52. Feise G, Kogure K, Busto R, Scheinberg P, Reinmuth OM: Effectof insulin hypoglycemia upon cerebral energy metabolism andEEG activity in the rat. Brain Res 126: 263-280, 1976

53. Wieloch T, Harris RJ, Symon L, Siesjd BK: Influence of severehypoglycemia on brain extracellular calcium and potassium ac-tivities, energy and phospholipid metabolism. J Neurochem 43:160-168, 1984

54. Hinzen DH, Becker P, Muller U: Einfluss von Insulin auf denregionalen Phospholipidstoffwechsel des Kaninchengehims invivo. Pflugers Archiv 321: 1-14, 1970

55. Lewis LD, Ljunggren B, Ratcheson RA, Siesjd BK: Cerebralenergy state in insulin-induced hypoglycemia, related to bloodglucose and to EEG. J Neurochem 23: 673-679, 1974

56. Ljunggren B, Schutz H, Siesjd BK: Changes in energy state andacid-base parameters of the rat brain during complete compressionischemia. Brain Res 73: 277-289, 1974

57. Auer RN, Hall P, Ingvar M, Siesjd BK: Hypotension as a compli-cation of hypoglycemia leads to enhanced energy failure but noincrease in neuronal necrosis. Stroke 17: 442—450, 1986

58. Pelligrino D, Yokoyama H, Ingvar M, Siesjd BK: Mocerate arte-rial hypotension reduces cerebral cortical blood flow and en-hances cellular release of potassium in severe hypoglycemia. ActaPhysiol Scand 115: 511-513, 1982

59. Auer RN, Kalimo H, Olsson Y, Siesjd BK: The temporal evolu-tion of hypoglycemic brain damage, n. Light and electron micro-scopic findings in the rat hippocampus. Acta Neuropathol (Bert)67: 25-36, 1985

60. Siesjd BK, Ingvar M, Pelligrino D: Regional differences in vascu-lar auto-regulation in the rat brain in severe insulin-induced hy-poglycemia. J Cereb Blood Flow Metabol 3: 478-485, 1983

61. Hill D, Loe PSU, Theobald J, Waddell M: A central homeostaticmechanism in schizophrenia. J Ment Sci 97: 111-131, 1951

62. Ratcheson R, Blank AC, Ferrendelli JA: Regionally selectivemetabolic effects of hypoglycemia in brain. J Neurochem 36:1952-1958, 1981

63. Harris RJ, Wieloch T, Symon L, Siesjd BK: Cerebral extracellu-lar calcium activity in severe hypoglycemia: Relation to extracel-lular potassium and energy state. J Cereb Blood Flow Metabol 4:187-193, 1984

64. Astrup J, Norberg K: Potassium activity in cerebral cortex in ratsduring progressive severe hypoglycemia. Brain Res 103:418-423

65. Cravioto RO, Massieu H, Izquierdo JJ: Free amino acids in ratbrain during insulin shock. Proc Soc Exp Biol 78: 856-858, 1951

66. Dawson RMC: Cerebral amino acids in fluoracetate poisoned,anaesthetized and hypoglycemic rats. Biochim Biophys Acta 11:548-552, 1953

67. DeRopp RS, Snedeker EH: Effect of drugs on amino acid levels inthe rat brain. J Neurochem 7: 128-134, 1961

68. Gorell JM, Dolkhart PH, Ferrcndelli JA: Regional levels of glu-cose, amino acids, high energy phosphates, and cyclic nucleotidesin the central nervous system during hypoglycemic stupor andbehavioral recovery. J Neurochem 27: 1043-1049, 1976

69. Abood LG, Geiger A: Breakdown of proteins and lipids duringglucose-free perfusion of the cat's brain. Amer J Physiol 182:557-560, 1955

70. Samson FE Jr, Dahl DR, Dahl N, Himwich HE: Studies of thehypoglycemic brain: Amino acids, nucleic acids, total nitrogenand side-group ionization of proteins in cat brain during insulincoma. Arch Neurol Psychiat 81: 458-465, 1959

71. Knauff HG, Marx D, Mayer G: Das Verhalten der Proteine undder serin-und colaminhaltigen Phosphatide des Zentralnervensys-



ownloaded from


HYPOGLYCEMIC BRAIN DAMAGE/Auer ion

tems wShrend der Insulinhypoglykfimie. Hoppe-Seylers Z PhysiolChem 326: 227-234, 1961

72. Hoshino Y, Elliot KAC: Effects of various conditions on themovement of carbon atoms derived from the glucose into and outof protein in rat brain. Can J Biochem 48: 236-243, 1970

73. Kiessling M, Weigel K, Gartzen D, Kleihues P: Regional hetero-geneity of L-(3-3H) tyrosine incorporation into rat brain proteinsduring severe hypoglycemia. J Cereb Blood Flow Metabol 2:249-253, 1982

74. Kiessling M, Xie Y, Kleihues P: Regionally selective inhibition ofcerebral protein synthesis in the rat during hypoglycemia andrecovery. J Neurochem 43: 1507-1514, 1984

75. Heald PJ: The effect of metabolic inhibitors on respiration andglycolysis in electrically stimulated cerebral-cortex slices. Bio-chem J 55: 625-631, 1953

76. Sandberg M, Nystr&m B, Hamberger A: Metabolically derivedaspartate — Elevated extracellular levels in vivo in iodoacetatepoisoning. J Neurosci Res 13: 489-495, 1985

77. Crone C: Facilitated transfer of glucose from blood into braintissue. J Physiol 181: 103-113, 1965

78. Gilboe DD, Betz AL: Kinetics of glucose transport in the isolateddog brain. Amer J Physiol 219: 774-778, 1970

79. Gjedde A, Diemer NH: Double-tracer study of the fine regionalblood-brain glucose transfer in the rat by computer-assisted auto-radiography. J Cereb Blood How Metabol 5: 282-289, 1985

80. Agardh C-D, Siesjd BK: Hypoglycemic brain injury: Phospholi-pids, free fatty acids, and cyclic nucleotides in the cerebellum ofthe rat after 30 and 60 minutes of severe insulin-induced hypogly-cemia. J Cereb Blood Row Metabol 1: 267-275, 1981

81. LaManna JC, Harik SI: Regional comparisons of brain glucoseinflux. Brain Res 326: 299-305, 1985

82. Agardh C-D, Kalmo H, Olsson Y, SiesjS BK: Hypoglycemicbrain injury: Metabolic and structural findings in rat cerebellarcortex during profound insulin-induced hypoglycemia and in therecovery period following glucose administration. J Cereb BloodFlow Metabol 1: 71-84, 1981

83. Smith M-L, Auer RN, Siesjd BK: The density and distribution ofischemic brain injury in the rat after 2-10 minutes of forebrainischemia. Acta Neuropathol (Berl) 64: 319-332, 1984

84. NevanderG, IngvarM, Auer RN, Siesjd BK: Status epilepticus inwell oxygenated rats causes neuronal necrosis. Ann Neurol 18:281-290, 1985

85. Auer RN, Kalimo H, Olsson Y, Wieloch T: The dentate gyms inhypoglycemia. Pathology implicating excitotoxin-mediated neu-ronal necrosis. Acta Neuropathol (Berl) 67: 279-288, 1985

86. Kalimo H, Auer RN, Siesjd BK: The temporal evolution of hy-poglycemic brain damage. HI. Light and electron microscopicfindings in the rat caudoputamen. Acta Neuropathol (Bert) 67:37-50, 1985

87. McLennan H, Collinridge GL, Kehl SJ: Electrophysiologkal ac-tions of kainate and other excitatory amino acids, and the structureof their receptors. In: Fuxe K, Roberts P, Schwarcz R (eds).Excitotoxins, Vol 39. MacMillan Press Ltd., London, pp 19-32,1983

88. Wieloch T: Hypoglycemia-induced neuronal damage preventedby an N-methyl-D-aspartate antagonist. Science 230: 681-683,1985

89. Wieloch T, Engelsen B, Westerberg E, Auer R: Lesions of theglutamatergic cortico-striatal projections ameliorate hypoglyce-mic brain damage in the striatum. Neurosci Lett 58: 25-30, 1985

90. Lindvall O, Auer RN, Siesjd BK: Selective lesions of mesostriataldopamine neurons ameliorate hypoglycemic damage in the cau-date-putamen. Exp Brain Res (in press), 1986

91. Wieloch T: The amino acid antagonist 2-amino-7-phosphono-heptanoic acid mitigates neuronal damage in the striatum due tohypoglycemia. Acta Physiol Scand (Suppl 542) 124: 397, 1985

92. Cserr HF, Cooper DN, Suri PK, Patlak CS: Efflux of radiolabeledpolyethylene glycols and albumin from rat brain. Am J Physiol240: F319-F328, 1981

93. Stern J, Hochwald GM, Wald A, Gandhi M: Visualization ofbrain interstitial fluid movement during osmotic disequilibrium.Exp Eye Res 25 (Suppl): 475-482, 1977

94. Klatzo I, Wisniewski H, Steinwall O, Streicber E: Dynamics of

cold injury edema. In: Klatzo I, Seitelberger F (eds). Brain Ede-ma. Springer-Verlag, New York, pp 554-563, 1967

95. Griffiths T, Evans MC, Meldrum BS: Temporal lobe epilepsy,excitotoxins and the mechanism of selective neuronal loss. In:Fuxe K, Roberts P, Schwarcz R (eds). Excitotoxins, Wenner-Gren International Symposium Series, Vol 39. Macmillan PressLtd., London, pp 331-332, 1983

96. Zanotto L, Heinemann U: Aspartate and glutamate induced reduc-tions in extracellular free calcium and sodium concentration inarea CA1 of 'in vitro' hippocampal slices of rats. Neurosci Lett35: 79-84, 1983

97. Glutamate neurotoxicity in cortical cell culture is calcium depen-dent. Neurosci Lett 58: 293-297, 1985

98. Wieloch T, Siesjd BK: Ischemic brain injury: The importance ofcalcium, lipolvtic activities, and free fatty acids. Path Biol 30:269-277, 1982

99. Brierley JB, Brown AW: Meldrum BS: The neuropathology ofinsulin induced hvpoglycemia in a primate (M. mullata); topogra-phy and cellular nature. In: Brierley JB, Meldrum BS (eds). BrainHypoxia. William Heinemann Medical Books Ltd., London. Ch22, pp 225-230, 1971

100. Kahn KJ, Myers RE: Insulin-induced hypoglycemia in the non-human primate. I. Clinical consequences. In: Brierley JB, Mel-drum BS (eds). Brain Hypoxia. William Heinemann MedicalBooks Ltd., London. Ch 19, pp 185-194, 1971

101. Meldrum BS, Horton RW, Brierley JB: Insulin-induced hypogly-cemia in the primate: Relationship between physiological changesand neuropathology. In: Brierley JB, Meldrum BS (eds). BrainHypoxia. William Heinemann Medical Books Ltd., London. Ch21, pp 207-224, 1971

102. Myers RE, Kahn KJ: Insulin-induced hvpoglycemia in the non-human primate, n. Long-term neuropathological consequences.In: Brierley JB, Meldrum BS (eds). Brain Hypoxia. WilliamHeinemann Medical Books Ltd., London. Ch 20, pp 195-206,1971

103. Nadler JV, Evenson DA, Cuthbertson GJ: Comparative toxicityof kainic acid and other acidic amino acids toward rat hippocam-pal neurons. Neuroscience 6: 2505-2517, 1981

104. Schwarcz R, Whetsell WO Jr, Mangano RM: Quinolinic acid: Anendogenous metabolite that produces axon-sparing lesions in ratbrain. Science 219: 316-318, 1983

105. Schwarcz R, Whetsell WO Jr, Foster AC: The neurodegenerativeproperties of intracerebral quinolinic acid and its structural analogcis-2,3-piperidine dicarboxylic acid. In: Fuxe K, Roberts P,Schwarcz R (eds). Excitotoxins, Vol 39. MacMillan Press Ltd.,London, pp 122-137, 1983

106. Vannucci RC, Duffy TE: Carbohydrate metabolism in fetal andneonatal rat brain during anoxia and recovery. Am J Physiol 230:1269-1275, 1976

107. HellmannJ, Vannucci RC, NardisEE: Blood-brain barrier perme-ability to lactic acid in the newborn dog: Lactate as a cerebralmetabolic fuel. Pediatr Res 16: 40-44, 1982

108. Queiroz L de Sousa, Eduardo RMP: Occurrence of dark neuronsin living mechanically injured rat neocortex. Acta Neuropathol(Berl) 38: 45-48, 1977

109. Ebels EJ: Dark neurons. A significant artifact: The influence ofthe maturational state of neurons on the occurrence of the phe-nomenon. Acta Neuropathol (Berl) 33: 271-273, 1975

110. Haworth JC, McRae KN: The neurological and developmentaleffects of neonatal hypoglycemia: A follow-up of 22 cases. CanMed Assoc J 92: 861-865, 1965

111. Banker BQ: The neuropathological effects of anoxia and hypogly-cemia in the newborn. Dev Med Child Neurol 9: 544-550, 1967

112. Anderson JM, Milner RDG, Strich SJ: Effects of neonatal hypo-glycemia on the nervous system: A pathological study. J NeurolNeurosurg Psychiat 30: 295-310, 1967

113. Jones EL, Smith WT: Hypoglycaemic brain damage in the neona-tal rat. In: Brierley JB, Meldrum BS (eds). Brain Hypoxia. Wil-liam Heinemann Medical Books Ltd., London. Ch 23, pp231-241, 1971

114. Saad SF: Further Observations on the role of y-aminobutyric acidin insulin-induced hypoglycaemic convulsions. Europ J Pharma-col 17: 152-156, 1972

115. Roberts E, Hammerschlag R: Amino acid transmitters. In: Albers



ownloaded from


708 STROKE V O L 17, No 4, JULY-AUGUST 1986

RW, Siegel GJ, Katzman R, Agranoff BW (eds). Basic Neuro-chemistry. Little, Brown and Co., Boston, pp 131-168, 1972

116. Meyer JS, Portnoy HD: Localized cerebral hypoglycemia simu-lating stroke. Neurology 8: 601-614, 1958

117. Moersch FP, Kemohan JW: Hypoglycemia. Neurologic and neu-ropathologic studies. Acta Neural Psychiat 39: 242-257, 1938

118. Tannenberg J: Comparative experimental studies onsymptomatology and anatomical changes produced by anoxic andinsulin shock. Proc Soc Exp Biol Med 40: 94-%, 1939

119. Tom MI, Richardson JC: Hypoglycemia from islet cell tumor ofpancreas with amyotrophy and cerebrospinal nerve cell changes. JNeuropathol Exp Neurol l(h 57-66, 1951

120. Winkelman NW, Moore MT: Neurohistopathologic changes withmetrazol and insulin shock therapy. Arch Neurol Psychiat 43:1108-1137, 1940

121. Silfverskiald BP: "Polyneuritis Hypoglycemia". Late peripheralparesis after hypoglycemic attacks in two insulinoma patients.Acta Med Scand 125: 502-504, 1946

122. Davies J, Evans RH, Jones AW, Mewett KN, Smith DAS, Wat-kins J: Recent advances in the pharmacoloy of excitatory aminoacids in the mammalian central nervous system. In: Fuxe K,Roberts P, Schwarcz R (eds). Excitotoxins, Vol 39. MacMillanPress Ltd., London, pp 43-54, 1983

123. Vital a , Picard J, Arne L, Aubertin J, Fenelon J, Mouton L:Pathological study of three cases of hypoglycemic encephalop-athy (one of which occurred after sulfamidotherapy). Le Diabete12F: 291-296, 1967

124. Kali mo H, Olsson Y: Effect of severe hypoglycemia on the humanbrain. Acta Neurol Scand 62: 345-356, 1980

125. Bodechtel G: Der Hypoglykamische Shock und seine Wirkungauf das Zentralnervensystem. Deutscbes Arch f klin Med 175:188-201, 1933

126. Lawrence RD, Meyer R, Nevin S: The pathological changes in thebrain in fatal hypoglycemia. Quart J Med 11: 181-201, 1942

127. Schleussing H, Schumacher H: Grosshirnschadigung im Verlauf

eines Diabetes mellitus. Dtsch Arch Klin Med 176: 45-51, 1933128. Carnmermeyer J: Uber Gehirnveranderungen, entstanden unter

Sakelscher Insulintherapie bei einem Schizophrenen. Z GesNeurol Psychiat 163: 617-633, 1938

129. Ddring G: Zur histopathologie and pathogenese des tddlichenInsuUnshocks. Deutsche Ztschr Nervenh 147: 217-227, 1938

130. Kastein GW: Insulinvergiftung. n. Neurologische und anato-misch-histologische Beschreibung. Z Ges Neurol Psychiat 163:342-361, 1938

131. Abdul-Rahman A,.Agardh C-D, Siesjd BK: Local cerebral bloodflow in the rat during severe hypoglycemia and in the recoveryperiod following glucose injection. Acta Physiol Scand 109:307-314, 1980

132. Myers RE: Lactic acid accumulation as a cause of brain edema andcerebral necrosis resulting from oxygen deprivation. In: KorobkinR, Guillenmineault G (eds). Advances in Perinatal Neurology.Spectrum Publishers, New York, pp 85-114, 1979

133. Auer RN, Ingvar M, Nevander G, Olsson Y, Siesjd BK: Earlyaxonal lesion and preserved microvasculature in epilepsy-inducednecrosis of the substantia nigra. Submitted for publication, 1986

134. Ingvar M, NevanderG, Siesjd BK: The development of hyperme-tabolic necrosis in the substantia nigra in status epilepticus. Sub-mitted for publication, 1986

135. Kuriyama M, Umezaki H, Fukuda Y, Osame M, Koike K, Tatei-shi J, Igata A: Mitochondrial encephalomyeloparhy with lactate-pyruvate elevation and brain infarctions. Neurology 34: 72-77,1984

136. Riggs JE, Schochet SS Jr, Fakacej AV, Papadimitriou A, Di-Mauro S, Crosby TW, Gutmann L, Moxley RT ID: Mitochondrialencephalomyopathy with decreased succinate cytochrome c re-ductase activity. Neurology 34: 48-53, 1984

137. Agardh C-D, Rosin I: Neurophysiological recovery after hypo-glycemic coma in the rat: Correlation with cerebral metabolism. JCereb Blood Flow Metab 3: 78-85, 1983



ownloaded from


R N AuerProgress review: hypoglycemic brain damage.

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1986 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.17.4.699

1986;17:699-708Stroke.

http://stroke.ahajournals.org/content/17/4/699World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Stroke Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:



ownloaded from

http://stroke.ahajournals.org/content/17/4/699

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://stroke.ahajournals.org//subscriptions/


progress review: hypoglycemic brain damag · progress review: hypoglycemic brain damag699 e roland...

Documents