Journal of the Neurological Sck'nces. 1113( 1991 ) $3-S 14
$3
© 1991 Elsevier Science Publishers B.V. 11022-5IOX/91/$03.50 JNS 03586
Traumatic brain injuries: structural changes J. Cerv6s-Navarro and J.V. Lafuente Inst. of Neuropathology, Free Unicersity Berlitz (Germany), and Dept. of Neuroseiences, Unirersity of Baske Country, Leioa, Vizcaya (Spain)
Summary A host of complications and consequences may follow a contusion or other brain injury of any sort. An appreciation of the temporal evolution of the contusion from a microscopic standpoint is useful to a full understanding of the process by which physical force damages the brain and how the brain reacts to this damage. Some disruptions of the blood brain barrier quite early will result in extraceilular edema. The microscopic appearance of an edematous area is usually spongy with numerous vacuoles. The neuropil may appear bubbly, and glial cells may be swollen. If edema has been long standing, the vacuoles may be larger and in fact a small cyst may appear in the white matter, if focal cerebral edema is not present for long periods of time and the underlying cause has been corrected, residual fluid and electrolytes are eventually removed, restoring the neuropii to a normal state, leaving no sign of its presence. However, in longer standing lesions, myelin pallor and some reactive gliosis may remain indefinitely. Neurons may show swelling very early and for a short period of time, which gives way to shrinkage, eosinophilia, and nuclear pyknosis. These changes may be observed at the periphery of lesions for as long as 5 or 6 months after the initial event. Before dissolution, nuclear pyknosis may remain in the tissue for many days and possibly longer, and may even become mineralized in situ (ferruginated neurons) to remain for years. In a traumatic lesion, swollen and ballooned axons may be found in and around the contusion but also at great distances from it (diffuse axonai injury). Axonai ballooning may be observed between 24 and 48 h postinjury and may persist wherever found for many years. Selective axonal calcification has been observed in humans as well as in experimental trauma. At about 7-10 days postinjury increased numbers of astroglia probably are present. Over the ensuing weeks and months, and probably years, astrocytes increase in number and in fibrillary appearance, eventually resulting in a glial scar in and about the injured area. It is thought that this reactive gliosis results in restoration of "the blood-brain barrier in the damaged area.
Introduction Traumatic brain injuries are a function of many distinct, though interactive phenomena such as the age of the victim, whether the head was moving at the time of impact or still relative to the impacting force, the location Of the impact site, the dimensions (area) of the impact site, whether or not the scalp a n d / o r skull was intact, the mass, velocity, and direction of the impacting event, the nature and extent of other injuries, as well as the physiological state of the individual at the time of the impact. Also important are a host of postimpact events, such as vital function status over time, the degree of hypoxemia or cerebral ischemia, changes in intracranial pressure, the presence or absence of subarachnoid, intraventricular, or intracerebral hemorrhage, whether operative intervention occured, and the duration of the postimpact clinical course prior the death. Primary brain damage by definition occurs at the moment of injury and could therefore be thought of as being irreversible. A host of complications and consequences may fol-
low a contusion or other brain injury of any sort. These t~ke the form of secondary lesions in the immediate vicinity of a contusion, global or more diffuse primary and secondary lesions usually in the white matter. Contusions are a type of focal brain damage brought about mainly by contact phenomena when the surface of the brain impacts on bony protuberances in the base of the skull. They have a highly characteristic appearance and distribution (Adams et al. 1980). They occur principally at the crests of gyri and in the acute stages are haemorrhagic. Contusions are most severe in the frontal and in the temporal lobes (Figs. 1 and 2). Contusions not obvious on external examination of the brain are frequently found on sectioning in the cortex above and below the sylvian fissure (Adams et al. 1986). An appreciation of the temporal evolution of the contusion from a microscopic standpoint is useful to a full understanding of the process by which physical force damages the brain and how the brain reacts to this damage (Oppenheimer 1968). The time course of reactions to cerebral traumatic contusion has been studied over the years but in a
$4 Gr~evic 1982), deep white matter pallor, and sometimes hydrocephalus.
General reactions
Fig. 1. Contusions on the surface of the frontal lobes =xtendinginto the adjacent white matter in a patient who surviveda few hours after a head injury. piecemeal manner. Much attention has been paid to early reactions, such as traumatically induced edema (Nevin 1967) and to late reactions, specifically in relation to the evolution and development of the macrophage (microglial) response (Oehmichen 1978), the evolution of the gliai scar (Clark 1974), and the fate of blood pigments (Kennady 1967). Comparatively few reports exist on an over-aU view of the traumatic process in the brain and its time course (Krauland 1973; Eisenmenger et al. 1978; Oehmichen 1978, 1980). Careful examinations of the brains of recovered posttrauma victims have shown a panoply of lesions in their brains that include evidence of axonai damage in the cerebral and brainstem white matter (axonal balloons seen in Bodian silver stained preparations) (Povlishock et al. 1983), focal astroglial scars (so-called glial nodules or inflammatory nodules) (Clark 1974;
Fig. 2. Old contusions in the frontal and temporal lobes in a patient who had made a good recoveryfrom a head injury.
Immediately after trauma, the blood brain barrier breaks down at the site of the injury. Generally, local vasculature is disrupted and blood cells and serum proteins massively invade the injury zone. About 24 h postlesion CNS edema is patent (Fig. 3). Edema probably results from both extracellular fluid accumulation, particularly in white matter, and astrocytic swelling. Abnormalities in the morphology of axons in both gray and white matter are observed almost immediatly after contusion injury, and uniform necrosis and myelin degeneration follows 8-24 h later. Accumulation of blood-derived macrophages are observed 48 h after a brain stab wound for example and these proceed to engulf degenerating myelin and other cell debris. Neurons not directly affected by injury begin to die 2 days postlesion and gradually enlarging cysts develop and coalesce to form cavities within the CNS parenchyma. Secondary neuronal death following the primary injury is probably responsible for a larger proportion of neuronal death than the initial injury itself. While phagocytosis is going on, astrocytes adjacent to the lesion proliferate and their enlarged fibrous processes form a web that isolates the surfaces of the injury from the surrounding tissue.
Early events
Edema The process of brain swelling or cerebral edema is probably the most common reaction to injury in the
Fig. 3. Cerebral edema in a patient who survived 24 h after a head injury.
$5 brain and is seen in so many contexts that, with the exception of hypoxic injury to the brain, it is the most constant accompaniment to every other form of injury. Some disruptions of the blood brain barrier quite early will result in extraceilular edema (Povlishock et al. 1978). if this edema is not compensable its mass and pressure effects may affect vascular perfusion in the region, resulting in secondary ischemia which will increase the edema. Regardless of which form of edema exists in response to trauma, it may spread weft outside the traumatized region and involve the whole hemisphere or the whole brain. In this instance there is a grave risk of excessive and irreversible increase in intracranial pressure, herniation and pertusion failure which may lead to brain death. Edema can occur in connection with contusions, lacerations, foreign bodies, or hemorrhages in the brain. In cases where there is a sudden subarachnoid hemorrhage, the irritating effect of blood on the external walls of brain vessels may prompt an edematous reaction which again acts to compound the mass effect of the blood, and hasten a fatal outcome. It is not known why this occurs, but clearly physical disruption of the capillary-brain barrier is likely, as well as chemical effects on the barriers by electrolytes such as calcium and potassium and inflammatory, mediators such as those released in the course of injury. lschemic injury results not only from direct vascular infarction but also from perturbation in the vasculature of a mechanically traumatized area. Reduction in local cerebral blood flow correlates with local anoxic damage of the brain vascular endothelium, an increase in the permeability of the blood brain barrier and local edema (Petito et al. 1982). Edema develops within 24 h from time of injury and is restricted to the white matter. These findings and the concomitant reduction in spatial buffering capacity suggests that the development of brain edema must be studied if we are to gain further insight into the pathophysiology of head injury. Traumatically induced edema is detectable within minutes of the injury, increasing over the next several hours, remaining stable for a few days, and decreasing and disappearing by about 6 days postinjury. Such edema, from whatever cause, is always capable of expanding the intrinsic mass effects of any lesion present in the adult, but in the child who has suffered head trauma cerebral edema may occur in connection with even apparently mild injury, and may lead to death (Pickles 1950; Langfitt et al. 1966; Zimmerman et ai. 1978; Barlow et al. 1983). A typical example of this phenomenon may be seen in the child who falls from a window to the pavement or is struck by a vehicle, who may or may not suffer a skull fracture and may or may not suffer loss of consciousness, but within hours of the traumatic episode may become stuporous
and drift into coma from elevated intracranial pressure. This phenomenon has been observed for many years by emergency room physicians and neurosurgeons, but has not been satisfactorily explained (Kobrine et al. 1977; Bruce 1982). A possible explanation may lie in the fact that cerebral blood flow in response to injury differs with age. in children under the age of 5 years, impacts to the head may result in increased cerebral blood flow, while in the adult the response may be just the opposite (Langfitt et al. 1966; Overgaard et al. 1974; Zimmerman et al. 1978). With increased blood flow into a possibly damaged vascular bed, the blood brain barrier may be more likely to open giving rise to massive edema which may or may not be fatal, Perhaps this special response and perhaps the greater "fragility" of the vascular bed of a child's brain is responsible for this peculiar phenomenon. The importance to the pathologist of this phenomenon is when attempting to develop a mechanism for death with the confusing and seemingly inconsistent finding of little evidence of brain trauma in the face of massive edema, with no obvious anatomic cause. In the adult, cerebral edema of a traumatic origin, if of long standing, may have deleterious effects on myelin which may lead to demyelination (Nevin 1967). Altered metabolic states such as prolonged acidosis, or associated conditions such as fat embolism, and the basic nature of shearing forces that might have injured long axons of passage can also be satisfactory explanations forapparent demyelination about contusions, or deeper brain lesions, but some workers feel that at least some of the demyelination observed in some cases is due to edema alone. This contention is difficult to prove or disprove. When edema is confined within cells as in so-called dry edema the process is most often diffuse rather than localized. In such a case, which is comparatively rare, the gross appearance of the brain may reflect its increased mass by showing a rounded and smooth surface where the gyri are flattened against the inner contour of the skull, and the sulci are compressed together. The brain will be increased in weight in the fresh state in proportion to the degree of edema, bearing in mind the normal weight ranges of brains in respect to race, age and sex. When the fixation takes place, the weight of the brain may fluctuate up or down during immersion in 10% formalin, depending upon the osmolality of the fixative. In general, after three weeks of fixation the brain will return very close to the weight in the fresh state. The process of fixation, except where osmolality of the fluid is very high, will not affect the appearance of any edema that may be present and will not obscure its pathology. On the cut section in dry edema, roundness and swelling of the brain are evident, as is a decrease in ventricular dimensions, if the process is diffuse, there
$6 will be no obvious asymmetrical midline shift of any structures. There may be hemorrhage or necrosis where uncal grooves or tonsillar grooves are especially severe, or where a vessel has been obstructed by the herniation. Although extensive dry edema may be observed from time to time it is usually associated with wet edema of the white matter in autopsy specimens. In these cases, after cutting the brain, fluid is either immediately apparent on the cut surface or gradually sweats from the surface to puddle in the hollows of the surface. In severe edema the white matter will have a creamy or rich yellow-green appearance and may even swell upon cutting. When there are localized lesions, which cause focal edema, the above changes may be confined to the immediate area of the lesion or may be diffuse and may even involve the opposite hemisphere though no lesion exists there. Histologically, depending on the duration and severity of the edema, the microscopic appearance of an edematous area is usually spongy with numerous vacuoles in the tissue (Fig. 4) which appear different from the usual vacuolating artifact of preparation commonly seen in blood vessels and nerve cells. The neuropil may appear bubbly, and glial cells may be swollen. Sometimes the process may appear more perivascular than diffuse, if edema has been long standing, the vacuoles may be larger and in fact small cysts may appear in the white matter. By employing immunohistochemical methods, which detect extravasated albumin or other serum proteins, the extent and severity of edema may be visualized in paraffin or frozen sections (Wilmes et al. 1979). Long-standing edema, of more than several weeks, may lead to reactive gliosis and the formation of larger, swollen, "gemistocytic" astrocytes in the affected region (Fig. 5). Because most edema fluid is rather low in protein content, the edematous tissue usually looks paler than normal white matter in H & E preparations.
q lie
L
:'+.'+ r',~+;I , ++.,+.,:++~:++t~+++,,
.,. , +.,+'+..~ ..'..++~ ,?',-. + ,.,~, ., , +.,++;,,.+.+Y+,
,,,~.:~" '.+!i~91' ~' + " " ' " + ++
, l .,'+'..e'.+.+'..,'+Pc + ;:+"++~:~; '+,.~'/, +,,~.~+/;.,'~
'+>'~+ " +, ;:~"+ , ~"'~':'~" i ~++"+l,.~TJ (,L+~+,!~..'~P + ¢ + :1+,"++.',~"+~." Nj~+.+,~j+~+.;~+:
"~' .....
,,'3
f.
,i+,,"' ~ , ' " . . . .
"~
+,i!~. + +.,_: .+.+.,
;.+0++ ,:, +,....,~.~,,..+.++++,,+,.~...+g:,,i~®r~j .+ +#.;~..,+..,~.+~+::~..~+++++~.+,,++ + +,++,.+:..+ •. . . . . . .,;.:. +.,,,. + + . +~,.+.,p,+ ~.~j.+~.g+-j~.:~ j,+9+ ~+.+ +.~+t-.~,"/'.'..,,~.+ ~/P',r+~I I % ~ .'+'`i~'+" ~.,+.",*, ~+
• ,+~'
"
~+,'+:,,.;'.
;'+.,'+¢.+!~+,"
+/+ : .
~11+ i
,,+ + .++;+.+ ++ +.l~q.,~j~+,~+~'"~+.p,,'++)+~.+~, ,+ +pc+'+ ,+, ~,, +,'~. t., + ~ ,+" +.+,+..+.,+++~+ , + . - , + i+~.~!+~?.?/++;++m+!: +-o +-+,.-,.~+'.,,'.-. .....• .... .+
,
,,++.. • +'
•
++~ +% ~,,,:.¢ + + / ; + ~+++.-'.
+,Pi,;~:.+~'. i
,";:+:~ +.+li~,+ +. ++++ ~:.' +~:,+~.' "++'+',.: ~'+i
+'
:~" I ~ ¢ ,+' ;~. + i`+ ~ , + d k l ~ ¢ "
' ++ ~+'+
.+;.,++': + +~,+;+i+.+.+~
~..-.'.~ • ++' + ~'~ .
,~,~,#P;V,~P,'~/mb,~,
+ ,J~i ++.+..~. I+ ;: , +, ' " +..+g~+,g+~f+?'+]e+.-+,++.+ +,:+.++~++;+++,,,r,+" "
., +" -
g+'~i'+ ,~+++. . . . . . ~+ + .+., . . . ., ., ., +,++ + ++~+.+,+ .~,+,+..,,~,.-.++..
•
.
+
/ ~+ .'~:;.+~+. spongiosus
in t h e w h i t e
+++.,.++++,
matter.
H&E,
" +, + .
_ + ,
,+, x 120.
F i g . 5. R e a c t i v e
astrocytes.
(a) Gallyas,
x 200, (b) H&E,
x 400.
"
' , +.;'J~:+++++....+,,.+.+,+:..+++++.,.++1+~+ +~ + + + ++:++ ~ #:+~" +++" + ~ ' : ? + ' , m". . . . . . +,++(.!+~+~,,~+.~+~,++~ +.++.,++.+~.++.~+:+ +..,. +,,.. .++-.:..++.+'~,.;+'~+/.':',"++)?i.~.~+~+.~,i+++++ +.+"I .',' +.+ ." • . , • .+ + 1 ~ +t" .+ " . "+ ~++.,~..+';'~:~; :+:.-+-....++!.,'..,, ++,. ~, ? . .+ ~.+ . ~+e .:,i~+. F i g . 4. S t a t u s
,,,
"0+~, +" ;+ ~' +'! ','.'++ +,+:+i t.'++ .++.+ ~+ '".+ ,+ +,...co +
,
$3
© 1991 Elsevier Science Publishers B.V. 11022-5IOX/91/$03.50 JNS 03586
Traumatic brain injuries: structural changes J. Cerv6s-Navarro and J.V. Lafuente Inst. of Neuropathology, Free Unicersity Berlitz (Germany), and Dept. of Neuroseiences, Unirersity of Baske Country, Leioa, Vizcaya (Spain)
Summary A host of complications and consequences may follow a contusion or other brain injury of any sort. An appreciation of the temporal evolution of the contusion from a microscopic standpoint is useful to a full understanding of the process by which physical force damages the brain and how the brain reacts to this damage. Some disruptions of the blood brain barrier quite early will result in extraceilular edema. The microscopic appearance of an edematous area is usually spongy with numerous vacuoles. The neuropil may appear bubbly, and glial cells may be swollen. If edema has been long standing, the vacuoles may be larger and in fact a small cyst may appear in the white matter, if focal cerebral edema is not present for long periods of time and the underlying cause has been corrected, residual fluid and electrolytes are eventually removed, restoring the neuropii to a normal state, leaving no sign of its presence. However, in longer standing lesions, myelin pallor and some reactive gliosis may remain indefinitely. Neurons may show swelling very early and for a short period of time, which gives way to shrinkage, eosinophilia, and nuclear pyknosis. These changes may be observed at the periphery of lesions for as long as 5 or 6 months after the initial event. Before dissolution, nuclear pyknosis may remain in the tissue for many days and possibly longer, and may even become mineralized in situ (ferruginated neurons) to remain for years. In a traumatic lesion, swollen and ballooned axons may be found in and around the contusion but also at great distances from it (diffuse axonai injury). Axonai ballooning may be observed between 24 and 48 h postinjury and may persist wherever found for many years. Selective axonal calcification has been observed in humans as well as in experimental trauma. At about 7-10 days postinjury increased numbers of astroglia probably are present. Over the ensuing weeks and months, and probably years, astrocytes increase in number and in fibrillary appearance, eventually resulting in a glial scar in and about the injured area. It is thought that this reactive gliosis results in restoration of "the blood-brain barrier in the damaged area.
Introduction Traumatic brain injuries are a function of many distinct, though interactive phenomena such as the age of the victim, whether the head was moving at the time of impact or still relative to the impacting force, the location Of the impact site, the dimensions (area) of the impact site, whether or not the scalp a n d / o r skull was intact, the mass, velocity, and direction of the impacting event, the nature and extent of other injuries, as well as the physiological state of the individual at the time of the impact. Also important are a host of postimpact events, such as vital function status over time, the degree of hypoxemia or cerebral ischemia, changes in intracranial pressure, the presence or absence of subarachnoid, intraventricular, or intracerebral hemorrhage, whether operative intervention occured, and the duration of the postimpact clinical course prior the death. Primary brain damage by definition occurs at the moment of injury and could therefore be thought of as being irreversible. A host of complications and consequences may fol-
low a contusion or other brain injury of any sort. These t~ke the form of secondary lesions in the immediate vicinity of a contusion, global or more diffuse primary and secondary lesions usually in the white matter. Contusions are a type of focal brain damage brought about mainly by contact phenomena when the surface of the brain impacts on bony protuberances in the base of the skull. They have a highly characteristic appearance and distribution (Adams et al. 1980). They occur principally at the crests of gyri and in the acute stages are haemorrhagic. Contusions are most severe in the frontal and in the temporal lobes (Figs. 1 and 2). Contusions not obvious on external examination of the brain are frequently found on sectioning in the cortex above and below the sylvian fissure (Adams et al. 1986). An appreciation of the temporal evolution of the contusion from a microscopic standpoint is useful to a full understanding of the process by which physical force damages the brain and how the brain reacts to this damage (Oppenheimer 1968). The time course of reactions to cerebral traumatic contusion has been studied over the years but in a
$4 Gr~evic 1982), deep white matter pallor, and sometimes hydrocephalus.
General reactions
Fig. 1. Contusions on the surface of the frontal lobes =xtendinginto the adjacent white matter in a patient who surviveda few hours after a head injury. piecemeal manner. Much attention has been paid to early reactions, such as traumatically induced edema (Nevin 1967) and to late reactions, specifically in relation to the evolution and development of the macrophage (microglial) response (Oehmichen 1978), the evolution of the gliai scar (Clark 1974), and the fate of blood pigments (Kennady 1967). Comparatively few reports exist on an over-aU view of the traumatic process in the brain and its time course (Krauland 1973; Eisenmenger et al. 1978; Oehmichen 1978, 1980). Careful examinations of the brains of recovered posttrauma victims have shown a panoply of lesions in their brains that include evidence of axonai damage in the cerebral and brainstem white matter (axonal balloons seen in Bodian silver stained preparations) (Povlishock et al. 1983), focal astroglial scars (so-called glial nodules or inflammatory nodules) (Clark 1974;
Fig. 2. Old contusions in the frontal and temporal lobes in a patient who had made a good recoveryfrom a head injury.
Immediately after trauma, the blood brain barrier breaks down at the site of the injury. Generally, local vasculature is disrupted and blood cells and serum proteins massively invade the injury zone. About 24 h postlesion CNS edema is patent (Fig. 3). Edema probably results from both extracellular fluid accumulation, particularly in white matter, and astrocytic swelling. Abnormalities in the morphology of axons in both gray and white matter are observed almost immediatly after contusion injury, and uniform necrosis and myelin degeneration follows 8-24 h later. Accumulation of blood-derived macrophages are observed 48 h after a brain stab wound for example and these proceed to engulf degenerating myelin and other cell debris. Neurons not directly affected by injury begin to die 2 days postlesion and gradually enlarging cysts develop and coalesce to form cavities within the CNS parenchyma. Secondary neuronal death following the primary injury is probably responsible for a larger proportion of neuronal death than the initial injury itself. While phagocytosis is going on, astrocytes adjacent to the lesion proliferate and their enlarged fibrous processes form a web that isolates the surfaces of the injury from the surrounding tissue.
Early events
Edema The process of brain swelling or cerebral edema is probably the most common reaction to injury in the
Fig. 3. Cerebral edema in a patient who survived 24 h after a head injury.
$5 brain and is seen in so many contexts that, with the exception of hypoxic injury to the brain, it is the most constant accompaniment to every other form of injury. Some disruptions of the blood brain barrier quite early will result in extraceilular edema (Povlishock et al. 1978). if this edema is not compensable its mass and pressure effects may affect vascular perfusion in the region, resulting in secondary ischemia which will increase the edema. Regardless of which form of edema exists in response to trauma, it may spread weft outside the traumatized region and involve the whole hemisphere or the whole brain. In this instance there is a grave risk of excessive and irreversible increase in intracranial pressure, herniation and pertusion failure which may lead to brain death. Edema can occur in connection with contusions, lacerations, foreign bodies, or hemorrhages in the brain. In cases where there is a sudden subarachnoid hemorrhage, the irritating effect of blood on the external walls of brain vessels may prompt an edematous reaction which again acts to compound the mass effect of the blood, and hasten a fatal outcome. It is not known why this occurs, but clearly physical disruption of the capillary-brain barrier is likely, as well as chemical effects on the barriers by electrolytes such as calcium and potassium and inflammatory, mediators such as those released in the course of injury. lschemic injury results not only from direct vascular infarction but also from perturbation in the vasculature of a mechanically traumatized area. Reduction in local cerebral blood flow correlates with local anoxic damage of the brain vascular endothelium, an increase in the permeability of the blood brain barrier and local edema (Petito et al. 1982). Edema develops within 24 h from time of injury and is restricted to the white matter. These findings and the concomitant reduction in spatial buffering capacity suggests that the development of brain edema must be studied if we are to gain further insight into the pathophysiology of head injury. Traumatically induced edema is detectable within minutes of the injury, increasing over the next several hours, remaining stable for a few days, and decreasing and disappearing by about 6 days postinjury. Such edema, from whatever cause, is always capable of expanding the intrinsic mass effects of any lesion present in the adult, but in the child who has suffered head trauma cerebral edema may occur in connection with even apparently mild injury, and may lead to death (Pickles 1950; Langfitt et al. 1966; Zimmerman et ai. 1978; Barlow et al. 1983). A typical example of this phenomenon may be seen in the child who falls from a window to the pavement or is struck by a vehicle, who may or may not suffer a skull fracture and may or may not suffer loss of consciousness, but within hours of the traumatic episode may become stuporous
and drift into coma from elevated intracranial pressure. This phenomenon has been observed for many years by emergency room physicians and neurosurgeons, but has not been satisfactorily explained (Kobrine et al. 1977; Bruce 1982). A possible explanation may lie in the fact that cerebral blood flow in response to injury differs with age. in children under the age of 5 years, impacts to the head may result in increased cerebral blood flow, while in the adult the response may be just the opposite (Langfitt et al. 1966; Overgaard et al. 1974; Zimmerman et al. 1978). With increased blood flow into a possibly damaged vascular bed, the blood brain barrier may be more likely to open giving rise to massive edema which may or may not be fatal, Perhaps this special response and perhaps the greater "fragility" of the vascular bed of a child's brain is responsible for this peculiar phenomenon. The importance to the pathologist of this phenomenon is when attempting to develop a mechanism for death with the confusing and seemingly inconsistent finding of little evidence of brain trauma in the face of massive edema, with no obvious anatomic cause. In the adult, cerebral edema of a traumatic origin, if of long standing, may have deleterious effects on myelin which may lead to demyelination (Nevin 1967). Altered metabolic states such as prolonged acidosis, or associated conditions such as fat embolism, and the basic nature of shearing forces that might have injured long axons of passage can also be satisfactory explanations forapparent demyelination about contusions, or deeper brain lesions, but some workers feel that at least some of the demyelination observed in some cases is due to edema alone. This contention is difficult to prove or disprove. When edema is confined within cells as in so-called dry edema the process is most often diffuse rather than localized. In such a case, which is comparatively rare, the gross appearance of the brain may reflect its increased mass by showing a rounded and smooth surface where the gyri are flattened against the inner contour of the skull, and the sulci are compressed together. The brain will be increased in weight in the fresh state in proportion to the degree of edema, bearing in mind the normal weight ranges of brains in respect to race, age and sex. When the fixation takes place, the weight of the brain may fluctuate up or down during immersion in 10% formalin, depending upon the osmolality of the fixative. In general, after three weeks of fixation the brain will return very close to the weight in the fresh state. The process of fixation, except where osmolality of the fluid is very high, will not affect the appearance of any edema that may be present and will not obscure its pathology. On the cut section in dry edema, roundness and swelling of the brain are evident, as is a decrease in ventricular dimensions, if the process is diffuse, there
$6 will be no obvious asymmetrical midline shift of any structures. There may be hemorrhage or necrosis where uncal grooves or tonsillar grooves are especially severe, or where a vessel has been obstructed by the herniation. Although extensive dry edema may be observed from time to time it is usually associated with wet edema of the white matter in autopsy specimens. In these cases, after cutting the brain, fluid is either immediately apparent on the cut surface or gradually sweats from the surface to puddle in the hollows of the surface. In severe edema the white matter will have a creamy or rich yellow-green appearance and may even swell upon cutting. When there are localized lesions, which cause focal edema, the above changes may be confined to the immediate area of the lesion or may be diffuse and may even involve the opposite hemisphere though no lesion exists there. Histologically, depending on the duration and severity of the edema, the microscopic appearance of an edematous area is usually spongy with numerous vacuoles in the tissue (Fig. 4) which appear different from the usual vacuolating artifact of preparation commonly seen in blood vessels and nerve cells. The neuropil may appear bubbly, and glial cells may be swollen. Sometimes the process may appear more perivascular than diffuse, if edema has been long standing, the vacuoles may be larger and in fact small cysts may appear in the white matter. By employing immunohistochemical methods, which detect extravasated albumin or other serum proteins, the extent and severity of edema may be visualized in paraffin or frozen sections (Wilmes et al. 1979). Long-standing edema, of more than several weeks, may lead to reactive gliosis and the formation of larger, swollen, "gemistocytic" astrocytes in the affected region (Fig. 5). Because most edema fluid is rather low in protein content, the edematous tissue usually looks paler than normal white matter in H & E preparations.
q lie
L
:'+.'+ r',~+;I , ++.,+.,:++~:++t~+++,,
.,. , +.,+'+..~ ..'..++~ ,?',-. + ,.,~, ., , +.,++;,,.+.+Y+,
,,,~.:~" '.+!i~91' ~' + " " ' " + ++
, l .,'+'..e'.+.+'..,'+Pc + ;:+"++~:~; '+,.~'/, +,,~.~+/;.,'~
'+>'~+ " +, ;:~"+ , ~"'~':'~" i ~++"+l,.~TJ (,L+~+,!~..'~P + ¢ + :1+,"++.',~"+~." Nj~+.+,~j+~+.;~+:
"~' .....
,,'3
f.
,i+,,"' ~ , ' " . . . .
"~
+,i!~. + +.,_: .+.+.,
;.+0++ ,:, +,....,~.~,,..+.++++,,+,.~...+g:,,i~®r~j .+ +#.;~..,+..,~.+~+::~..~+++++~.+,,++ + +,++,.+:..+ •. . . . . . .,;.:. +.,,,. + + . +~,.+.,p,+ ~.~j.+~.g+-j~.:~ j,+9+ ~+.+ +.~+t-.~,"/'.'..,,~.+ ~/P',r+~I I % ~ .'+'`i~'+" ~.,+.",*, ~+
• ,+~'
"
~+,'+:,,.;'.
;'+.,'+¢.+!~+,"
+/+ : .
~11+ i
,,+ + .++;+.+ ++ +.l~q.,~j~+,~+~'"~+.p,,'++)+~.+~, ,+ +pc+'+ ,+, ~,, +,'~. t., + ~ ,+" +.+,+..+.,+++~+ , + . - , + i+~.~!+~?.?/++;++m+!: +-o +-+,.-,.~+'.,,'.-. .....• .... .+
,
,,++.. • +'
•
++~ +% ~,,,:.¢ + + / ; + ~+++.-'.
+,Pi,;~:.+~'. i
,";:+:~ +.+li~,+ +. ++++ ~:.' +~:,+~.' "++'+',.: ~'+i
+'
:~" I ~ ¢ ,+' ;~. + i`+ ~ , + d k l ~ ¢ "
' ++ ~+'+
.+;.,++': + +~,+;+i+.+.+~
~..-.'.~ • ++' + ~'~ .
,~,~,#P;V,~P,'~/mb,~,
+ ,J~i ++.+..~. I+ ;: , +, ' " +..+g~+,g+~f+?'+]e+.-+,++.+ +,:+.++~++;+++,,,r,+" "
., +" -
g+'~i'+ ,~+++. . . . . . ~+ + .+., . . . ., ., ., +,++ + ++~+.+,+ .~,+,+..,,~,.-.++..
•
.
+
/ ~+ .'~:;.+~+. spongiosus
in t h e w h i t e
+++.,.++++,
matter.
H&E,
" +, + .
_ + ,
,+, x 120.
F i g . 5. R e a c t i v e
astrocytes.
(a) Gallyas,
x 200, (b) H&E,
x 400.
"
' , +.;'J~:+++++....+,,.+.+,+:..+++++.,.++1+~+ +~ + + + ++:++ ~ #:+~" +++" + ~ ' : ? + ' , m". . . . . . +,++(.!+~+~,,~+.~+~,++~ +.++.,++.+~.++.~+:+ +..,. +,,.. .++-.:..++.+'~,.;+'~+/.':',"++)?i.~.~+~+.~,i+++++ +.+"I .',' +.+ ." • . , • .+ + 1 ~ +t" .+ " . "+ ~++.,~..+';'~:~; :+:.-+-....++!.,'..,, ++,. ~, ? . .+ ~.+ . ~+e .:,i~+. F i g . 4. S t a t u s
,,,
"0+~, +" ;+ ~' +'! ','.'++ +,+:+i t.'++ .++.+ ~+ '".+ ,+ +,...co +
,
