In utero, neonates, infants and children william a. Cox, M. D. Forensic pathologist

SUBDURAL HEMATOMAS AND RELATED LESIONS

IN UTERO, NEONATES, INFANTS AND CHILDREN

WILLIAM A. COX, M.D.

FORENSIC PATHOLOGIST/NEUROPATHOLOGIST
November 24, 2009
Subdural hemorrhage is a common sequale in non-accidental trauma of the infant and

Child. In one study non-accidental trauma was responsible for acute subdural hemorrhage in 84% of infants and in 81% in older children. In another study non-accidental trauma was responsible for acute subdural hemorrhage in 82% of infants and children under the age of 2 years.

In Utero

Subdural bleeding in utero, including hematomas also occurs. The etiopathogenesis of intrauterine subdural hemorrhage (SDH) cannot always be defined. The fetus is usually well protected from direct abdominal trauma by maternal structures and amniotic fluid. Fetal intracranial vessels may be susceptible to shearing or acceleration/deceleration forces because of the following features: First, the head is large and the neck muscles are weak; this allows more rotational movement with angular acceleration. Second, the subarachnoid space is larger, allowing the brain to move within the cranium easily. Third, the fetal brain has a higher water content that increases its mass and allows if to develop more momentum when acceleration is applied. These unique conditions could make the fetus more susceptible to developing subdural hematomas without obvious trauma.

The type of hemorrhagic lesion a fetus will show depends greatly on their maturity. Hemorrhages in prematures are mainly subependymal and intraventricular, leptomeningeal, or in the cerebellar parenchyma, and are related to asphyxia. They become less frequent as the infant approaches maturity. The types of hemorrhages characteristic of the mature newborn are related to mechanical trauma, such as lacerations of tentorium, falx or large venous channels, subdural hemotomas, or injuries to the spinal cord. The history of the mode of delivery is of extreme importance and ventouse extraction is recognized to have an incidence of SDHs, most of which resolve after birth. MRI in routine deliveries identifies asymptomatic SDH in a proportion. These usually resolve within 1 month.

Before continuing with a discussion of subdural hemorrhages, there are hemorrhages, which involve the falx cerebri (intralfalcine) as well as the dura (intradural), which need to be discussed. Hemorrhages into the loose connective tissue of the immature falx or tentorium are quite common in term as well as the premature infant.

Typically the dura is looked upon as a tough fibrous covering of the brain, being relatively avascular with the exception of containing the dural sinuses. However, the dura in the fetal brain is a complex vascular and innervated structure rather than the fibrous avascular structure of the adult brain. The dura contains an inner vascular plexus that is larger in the infant than in the adult. It is believed this plexus plays a role in the absorption of cerebral spinal fluid in light of the fact the arachnoid granulations in the fetal brain are not completely developed. Although subdural hemorrhages in the fetus frequently have a traumatic causation, there are nontraumatic conditions associated with subdural hemorrhage, and the inner dural plexus is a likely source of bleeding in these nontraumatic circumstances.

Intrafalcine and intradural hemorrhages are usually multicentric and range from small petechiae to large, confluent effusions, commonly involving the falx, particularly in the portion near the superior sagittal sinus. Larger hematomas, of 2 to 3 mm thickness, may cause the flax or the tentorium to bulge, penetrating to the very surface of the tissue. It is uncommon for these intradural hemorrhages to rupture into the subdural space. They are of no clinical significance. They have been attributed to either asphyxia or mechanical trauma, but that has not been affirmed. Their resorption occurs very quickly and without difficulty; infants a few months old rarely show organizing hematomas or pigmentation in falx or tentorium despite the fact that they occur with great frequency in newborns.

Lacerations of the tentorium, however, are a far more serious lesion than intradural hemorrhages. The frequency of tentorial lacerations increases with gestational age or with birth weight. They are rare in prematures with two-thirds of the infants with tentorial laceration being in the weight group of over 4,500 grams, compared with an incidence of only 4 % in prematures.

When reviewing the older literature it becomes clear that the frequency of lacerations of the tentorium have decreased markedly with improved techniques of obstetric management. As an example, Holland (1921) reported tentorial lacerations in 48% of 168 autopsies, whereas Resbitt and Anderson (1956) reported a frequency of deaths from dural laceration of 1 in 2,060 births.

Lacerations of the tentorium occur most commonly at its free edge and may be complete or incomplete, affecting either the superior or the inferior leaf of the tentorial tissue. Ragged edges are seen in the torn tissue. Hemorrhages ensue from rupture of intratentorial blood vessels, and they attain major extent from the opening of larger venous channels. Extension of the tear in the median direction opens the straight sinus, while lateral extension opens the transverse sinus. The vein of Galen may be torn off its insertion at the straight sinus, but this type of injury is rather rare and is often confused with large asphyctic cerebellar hematomas. Tears of the sinuses give rise to massive subdural hemorrhages. Subdural hemorrhages from tearing of bridging veins from the saggital sinus is considered the most common cause, however, well documented cases are very difficult to find. Oval tears of the falx may be seen in association with tentorial tears; however, they can also be seen singularly. Such singular lacerations of the falx are much less frequent than in the tentorium comprising between 4 to 12% of all lacerations. One of the problems with these figures is that the falx often contains fenestration as a part of normal development, which some believe are misinterpreted for focal areas of disruption, i.e., lacerations, hence the 4 to 12% figure may be actually inflated. However infrequent, subdural hemorrhage associated with tentorial tears represent the classic form of cranial birth trauma.

The actual mechanism of injury to the veins and subdural folds appears to be the result of excessive fronto-occipital compression or from oblique distortion. The fetal head is well designed to withstand evenly applied compression, however, fronto-occipital compression may cause kinking, obstruction and tears in the Great cerebral vein (vein of Galen) or the closely related superior cerebellar veins. You may also see rupture of the longitudinal sinus due to the overlapping of the parietal and occipital bones. Oblique compression, such as may be caused by an incorrect application of forceps, places one tentorial leaf under tension while relaxing the opposite one; thus you may see tearing of the bridging veins by the overriding of parietal bones. In vaginal breech delivery it is not uncommon to find tentorial tears accompanying occipital osteodiastasis.

Occipital osteodiastasis (OOD) is a form of birth injury characterized by a tear along the innominate (posterior occipital or supraoccipital-exoccipital) synchondrosis with separation of the occipital squama from the lateral or condylar parts of the occipital bone. This condition is due to excessive pressure exerted over the subocciput during delivery, resulting in a forward and upward displacement of the anterior margin of the occipital squama into the posterior cranial fossa, giving rise to posterior fossa hemorrhage and other intracranial complications, which are invariably fatal. Although, this condition was reported in the older literature as being relatively common it is relatively uncommon today, which is most likely the result of improved obstetric techniques. There is a less severe form of OOD, which is survivable. OOD has also been reported in the postnatal period (this period begins immediately after birth and extends for about 6 weeks). OOD has been reported in a 3-month-old child with a diagnosis of child abuse who survived and a 2-year-old who was involved in a fatal motor-pedestrian collision.

In a general sense the following circumstances are the underlying causation of tentorial tears: Rapid forcing of the fetal head through the birth canal, such as in precipitate labor, breech presentation or extraction; a disproportion between the fetal head and the pelvis; cranial trauma caused by forceps; rigidity of the soft parts of the birth canal.

Subdural hemorrhage due to tears in the bridging veins and tears of the dural folds are commonly found together because they result from the same type of injury. In those cases in which the subdural hemorrhage is space occupying, giving rise to obvious mass effect it is a cause of death. However, a more frequent finding is the association of tears of one or both leaves of the tentorium with minimal subdural bleed. In such cases the question arises as to the severity of the cranial trauma and its relationship to the death. What must be understood is that tears of the dural folds can only occur as a result of gross distortion of the cranium. Such gross distortion of the cranium can be expected to give rise to trauma to the brain regardless of either absence of or only minimal subdural hemorrhage. Thus, if tears of the dural folds are identified, with or without subdural hemorrhage, they should be regarded as being a manifestation of severe trauma to the cranial vault and the underlying cause of death.

It is important that it is understood that not all hemorrhages from the great cerebral vein (Galen) are the result of trauma. Malformations of the vein of Galen, although usually presenting with congestive heart failure and a steal phenomenon, giving rise to areas of infarction of the brain, can also rupture giving rise to subarachnoid and subdural hemorrhage. Sometimes the vein of Galen malformation can undergo thrombosis resulting in a large, central, hemorrhagic infarction.

Although the majority of subdural and subarachnoid hemorrhages are attributed to traumatic deliveries, they can also be the result of hypoxic insults.

In those fetuses and neonates that survive, resorption of these hematomas is slower than that of subarachnoid hemorrhages because of their larger size and because of the inherently slower resorption of fluids from the subdural space. Large organizing hematomas have been found attached to the tentorium in infants several months old. Subdural hematomas are a potential source of chronic subdural hematomas or hygromas, especially those of the inferior cerebral fossa; these will be discussed later in this article.

Subdural hemorrhage in Neonates and Infants

Anatomically these lesions occur in two areas within the cranial vault, supratentorial and infratentorial. The supratentorial region of the brain is the area located above the tentorium cerebelli; the area of the brain below it is the infratentorial region. The supratentorial region contains the cerebral hemispheres, while the infratentorial region contains the cerebellar hemispheres.

The symptoms of acute supratentorial subdural hemorrhage at birth include a bulging fontanel, shock with pallor, papillary dilatation, twitching or seizures, vomiting, irritability and restlessness and stupor. Most of these infants are born at term, commonly with a history of traumatic or instrumental delivery. Subdural hematomas of postnatal onset commonly develop following craniocerebral trauma, often associated with linear skull fractures.

These hemorrhages are commonly bilateral, usually in the parieto-occipital region or the posterior aspect of the interhemispheric fissure. The underlying causation is the stretching and tearing of the bridging veins at their entrance into the superior sagittal sinus due to deformation of the head by traumatic birth or by postnatal craniocerebral trauma. The hemorrhages in the infant group typically consist of a thin rim of liquid blood or having the appearance of ‘current jelly.’ Due to their small quantity they seldom give rise to mass effect, thus rarely requiring neurosurgical intervention. In the older child the subdurals are generally more adult-like in that they are unilateral and space occupying.

Infratentorial subdural hematomas cause changes in respiration in terms of rate, depth or rhythm, changes in cry, vomiting, poor suck and general hypotonia. Subdural hematomas at this site are most likely caused by lacerations of the tentorium, which manifest by attachment of organized hematomas to the falx or the formation of neomembranes in association with an organized hematoma between the leaflets of the falx. Clinically identified subdural hematomas in the posterior fossa of newborns are rare; these lesions are much less common than supratentorial hematomas. The underlying reason these lesions are so seldom seen today in comparison to the older literature is due to the improvement in obstetrical management as well as correctly identifying asphyctic subarachnoid or cerebellar hemorrhages rather than calling such hemorrhages ‘lesions of birth trauma.’ Rarely subdural hematomas in the posterior fossa can give rise to upward herniation of the cerebellum through the tentorial hiatus causing compression and focal hemorrhage, necrosis and sclerosis of the cerebellar cortex at the edge of the tentorium; these changes were typically associated with tonsillar hemorrhage, necrosis and sclerosis due to herniation into the foramen magnum.

Although most subdural hemorrhages in infants and children have a traumatic basis, some do not. Inherited disorders of bleeding such as deficiencies of Factor I, II, V, VII (hemophilia A), IX (hemophilia B), X, SI and XII, von Willebrand’s Disease, Activated Protein C Resistance, Glanzmann thrombasthenia, Protein C deficiency, Wiscott-Aldrich Syndrome, Bernard-Soulier Syndrome, antithrombin III deficiency, and Hereditary deficiency of platelet storage granules (Dense body deficiency of Gray or Gray platelet syndrome). There are also acquired deficiencies, which can produce subdural hematomas such as vitamin K deficiency, liver disease, disseminated intravascular coagulation, and the development of circulating anticoagulants. Severe liver disease can give rise to impaired clotting factor synthesis or decreased hepatic synthesis of alpha2-antiplasmin.

There are other conditions, which can give rise to bleeding such as vascular malformations and metabolic disorders, e.g. glutaric acidemia type 1, which gives rise to thinning of the vessel walls which can result in spontaneous bleeding, hypernatremia, mitochondrial disorders such as Alpers disease and Menke disease. Infants with enlarged extra-axial spaces, such as in shunted hydrocephalus appear to be at increased risk of subdural bleeding with lesser degrees of trauma.

Subdural hemorrhage in the Child

In the child the subdural hemorrhages are generally more adult-like in that they are unilateral and space occupying. A skull fracture is found in approximately 30% of cases. They have a poor prognosis with mortality rates between 60 to 90%. If the patient has surgery within 4 hours of the initial injury this mortality rate drops to 30%. In actuality the significant mortality rate is due to the often-associated brain damage. Typically, subdural hematomas do not occur in isolation, but rather are a component of additional head injuries. The associated conditions include focal or diffuse cerebral edema, diffuse axonal injury, severe contusions, intraparenchymal hemorrhage, and epidural hematomas.

Clinically these patients present with a history of head trauma accompanied by loss of consciousness, with some recovery giving rise to a lucid interval, but not completely back to normal. It is during this lucid interval the patient may complain of a headache, show personality changes, confusion and drowsiness, lethargy, stiff neck, irritability, vomiting, seizures, papillary dilatation, and a low-grade fever.

These symptoms and signs can evolve over minutes, hours, days or longer. The length of time of the lucid interval is dependent upon the rate of accumulation of the blood and the ability of the brain to compensate for the expanding mass. What is also important as far as the length of time of the lucid interval is the coexistence of other traumatic lesions, including contusions, diffuse axonal injury, hypoxic-ischemic events, and intraparenchymal hemorrhage all of which are associated with the genesis of cerebral edema and increasing intracranial pressure (ICP). The coexistence of other existing mass lesions such as arachnoid cyst, subdural hygromas, subdural effusions, and hydrocephalus can further complicate the picture due to the fact they decrease the brains ability to compensate for the mass effect of the subdural. What is also important to remember is their symptoms may not progress in a steady linear fashion; instead their symptoms may actually appear to improve for a brief period of time only to then begin to worsen. Once the ICP reaches 20 mmHg the brain can no longer compensate. This ultimately culminates in the cerebral circulation being severely compromised leading to deepening unconsciousness, coma, respiratory depression and cardiac arrest.

Chronic Subdural Hemorrhage in Neonates, Infants and Children

What percentage of acute subdural hematomas in neonates, infants and children resolve spontaneously, whereas others go on to develop into chronic subdural hematomas is not known? It is clear that in some infants their acute subdural hematomas completely resolve never showing symptoms. Why some acute subdural hematomas develop into chronic subdural hematomas is not known, although some have suggested their evolution is due to the mixing of CSF and venous blood. This latter thought is based on the work of Watanbe and coworkers who were able to produce a chronic subdural hematoma by mixing blood with CSF. They found by mixing freshly collected canine whole venous blood and human or canine CSF in ratios of 20:1 to 5:1 and held at 37 degrees they could form a lesion analogous to a chronic subdural hematoma. Regardless of the underlying genesis of chronic subdural hematomas, they have a propensity to enlarge secondary to rebleeding. It is believed that the rebleeding is due to leakage from capillaries in the neomembranes of the chronic subdural. This bleeding can be spotty and intermittent over weeks to months. There are occasions in which the rebleeding can be substantive giving rise to sudden mass effect followed by decompensation and death. What is important to understand is that this rebleeding is often spontaneous and is not representative of new trauma.

There is another important complication of subdural hematomas that must be kept in mind and that is whether the subdural hematoma is acute, subacute or chronic, they can give rise to poorly perfused cortical areas directly beneath them. Should the child survive these poorly perfused cortical areas will undergo necrosis leading to focal areas of cortical atrophy and on occasion, cystic degeneration of the cortex and underlying white matter. Sometimes these areas of cortical and subcortical ischemia will extend beyond the area of the subdural hematoma. The underlying pathogenesis of these cortical and subcortical areas of ischemia is believed to be related to interference or compromise of the microcirculation in the arachnoid and on the cortical surface.

The development of chronic subdural hematomas in neonates and infants is manifested by lack of normal weight gain, irritability, poor feeding and, sometimes, an accelerated increases in head size. Cranial enlargement may be asymmetrical, and transillumination may occur if there is a hygroma filled with clear fluid. In one study, subdural hematomas were responsible of 9% of 310 babies with an enlarged head (Ingraham and Matson, 1944).

Infratentorial chronic subdural hematomas may give rise to hydrocephalus by obstructing cerebral spinal fluid flow. Hydrocephalus in infants also often coexists with large supratentorial chronic subdural hematomas or hygromas. Sometimes this particular combination is interpreted as posthemorrhagic hydrocephalus due to traumatic birth. However, there is very little evidence to show that supratentorial subdural hemorrhage or encapsulated subdural hematomas cause hydrocephalus. In point of fact the cause-effect relationship may actually be reverse, as shown by Anderson (1952) who observed the formation of unilateral or bilateral subdural hemorrhages in 3 hydrocephalic infants following a sudden collapse of the brain from the draining of intraventricular fluid by shunting or by intraventricular surgery. In these cases the stretching and tearing of the bridging veins near the sinus due to the retraction of the collapsing cerebral hemispheres evidently caused the subdural hemorrhage. Becker and Nulsen (1968) observed subdural hemorrhages as a complication of shunting in 5% of 140 hydrocephalic patients. The mechanism responsible for the formation of subdural hemorrhages is potentially repetitive, and may give rise to bilateral chronic subdural hematomas with many layers of neomembranes, or recent hemorrhages evolving into older fibrotic hygromas.

Chronic subdural hematomas may be found in infants having a variety of atrophic or destructive cortical lesions. Instances of unilateral subdural hematomas associated with unilateral ulegyria, or with unilateral progressive sclerosing cortical atrophy (hemiatrophy) have been reported. Christensen and Højgaard (1964) reported bilateral subdural hematomas associated with progressive sclerosing cortical atrophy; Griepentrog (1952) reported chronic subdural hematomas with hydranencephaly.

The shrinkage of atrophic cerebral hemispheres may result in the stretching of cortical bridging veins, which traverse the greatly enlarged subarachnoid space. The assumption that chronic subdural hematomas may be secondary to hemispheric atrophy may also pertain to the observation that subdural hematomas form as a complication of shunting procedures for “normal pressure hydrocephalus” in adults Isamuelson et al., 1972).

Physical force required to produce a Subdural Hematoma in an infant or child

It is not known what the lowest injury threshold for infantile subdural hematomas might be, but from all available evidence it appears that blunt impact injury, either due to a fall or an assault, that generate a force of 100 g or greater can produce an acute subdural hemorrhage in normal infants. Utilizing instrumented dummies, which were allowed to fall to a dense carpeted surface from 36" to 48" heights, in which the head struck the surface, peak g forces of 121 to 189 g were recorded with angular accelerations from 7,600 to 21,700 radians/second/second and velocities of 25 to 49 radians/second. In another experiment, utilizing dummies simulating a 3-year-old child, impact forces of 287 to 349 g were recorded with angular accelerations of 16, 000 to 79,000 radians/second/second and velocities of 16 to 65 radians/second. Thus, a fall of 36" to 48" onto a floor, regardless of its coverings, can generate an impact force of 100 g or greater giving rise to an acute subdural hemorrhage. Although the statistics clearly show the most common underlying cause for subdural hemorrhage in the infant and child is non-accidental trauma, a fall of as little as 36" can give rise to an acute subdural hemorrhage.

The expression radian per second squared is the SI unit of angular acceleration. It is the increase in the change in orientation in radians of an object per second per second; in other words, the increase in angular velocity (measured in radians per second) per second or the rate of change of angular velocity over time. The expression radian per second is the SI unit of angular velocity. The radian per second is defined as the change in orientation of an object, in radians, every second. It is also a unit of angular frequency. Angular frequency is the magnitude of the vector quantity angular velocity. A radian is a unit of plane angle. An angle of 1 radian results in an arc with an equal length to the radius of the circle.

There is another controversial issue that needs to be addressed and that is the relationship between ‘shaken baby-syndrome’ and the evolution of acute subdural hematomas. In 1974 Caffey coined the term ‘whiplash shaken infant syndrome.’ As originally used this term describes a constellation of injuries seen with regularity in infants who have been physically abused: these include diffuse brain injury with altered consciousness, subdural and subarachnoid hemorrhage over the convexities, retinal hemorrhages, and scalp contusions. Many authors also included metaphyseal avulsions in this constellation, as Caffey did. It has been strongly debated over the years as to whether violent shaking in of itself, without impact, can give rise to acute subdural hemorrhage.

Several experiments have been done examining the force parameters produced by shaking dummies that duplicate the physics occurring during shaking events. What these studies have clearly shown is that no matter what effort is applied in shaking a baby dummy the acceleration duration time for one pulse is about 10 times longer than that occurring in a fall-type impact (250 milliseconds vs. 20 milliseconds), and peak accelerations for the shaking are less than 25 g (average 10-12 g). These parameters are clearly below the known thresholds for both subdural hematomas and brain injury. Thus, shaking in of itself without associated impact cannot generate sufficient force to produce acute subdural hemorrhage and brain injury. Shaking can however produce soft tissue injury in the neck as well as cervical spinal cord injury. There is however a report in the literature of violent shaking during torture giving rise to acute SDH in adults, (Pounder et al., 1997).

Subdural Hygroma

The term subdural hygroma designates an encapsulated subdural lesion filled with clear or xanthochromic fluid. Various pathogenetic mechanisms have been proposed for the formation of these lesions. One line of thought is that hygromas are the residual cavity of a large, incompletely organized chronic subdural hematoma. Liquefaction of the hematoma transforms it into a fluid of low viscosity which, with the passage of time, turns from turbid dark brown to xanthochromic and finally to clear fluid. The capsule of collagenous tissue that had formed around the chronic subdural hematoma prevented it from collapsing during this entire process. This interpretation of hygromas is supported by the observation of hematomas in various stages of organization and transition to hygromas, and by microscopic evidence of old hemorrhage in the walls of hygromas.

There is another line of thought, which suggests that the formation of hygromas is the result of exudates of clear fluid into the subdural space. This fluid is due to the escape of CSF from the subarachnoid space through a defect in the arachnoid membrane due to craniocerebral trauma or hydrocephalus, or from a clear transudate caused by increased vascular or membrane permeability due to inflammation involving the leptomeninges. It is not feasible to distinguish these two modes of formation of subdural hygromas on the basis of morphology alone. However, subdural hygromas usually occur under conditions in which hematomas are found during the acute phase of the disease. This suggests that hemorrhage generally is a more important factor in hygroma formation than the leakage of CSF through a defect in the arachnoid or as the result of a clear transudate.

Traumatic Subdural Effusions

As indicated above, the formation of subdural hygromas has been attributed to a rupture of the arachnoid membrane in craniocerebral trauma, whereby a ball valve action results in hygroma formation. However, it is generally believed that it is unlikely that such lesions occur in the absence of subarachnoid or subdural hemorrhage. It is also believed that there are some instances in which localized fluid accumulation attributed to a hygroma is in reality an arachnoid cyst.

Subdural Effusions in Leptomeningitis

Subdural effusions can result from leptomeningitis, repeated subdural taps or excessive draining of CSF or from abnormal transudates. Neomembranes have been found frequently in association with these subdural effusions, but there is little evidence to suggest that they progress to form encapsulated hygromas. Subdural effusions of clear fluid may also result from the presence of an infective process in the epidural space, such as in otitis media.

Hyperacute Subdural Hematomas

On occasion subdural hematomas as well as epidural hematomas will be described as ‘hyperacute.’ This term has its foundation in the density with which subdural and epidural hematomas present on CT scan. The density of these hematomas is related to the attenuation values of the clot, as a function of the erythrocyte and hemoglobin protein concentration and to a lesser extent the iron content of the hemoglobin molecule. Blood itself has an attenuation value of 20-30 HU (Hounsfield units).

The Hounsfield scale is a quantitative scale for describing radiodensity. The Hounsfield unit (HU) scale is a linear transformation of the original linear attenuation coefficient measurement in one in which the radiodensity of distilled water at standard pressure and temperature (STP) is defined as zero Hounsfield units, while the radiodensity of air at STP is defined as –1000 HU. For a material X with linear attenuation coefficient μx, the corresponding HU value is therefore given by μx – μH2O divided by μH2O times 1000, where μH2O is the linear attenuation coefficient of water. Thus, a change of one Hounsfield unit represents a change of 0.1% of the attenuation coefficient of water since the attenuation coefficient of air is nearly zero. It is the definition for CT scanners that are calibrated with reference to water.

Clots are mainly composed of RBC, hemoglobin protein and fibrin, so their hematocrit level is higher than that of circulating blood and consequently, their attenuation is higher. Chronic clots have a very high attenuation value, typically in the range of 87 HU ± 30. Some believe the high attenuation of chronic clots is caused by the vascularization or the calcification of the clots themselves. Clearly in subdural hematomas, calcification is not an issue.

Serum hemoglobin concentrations ranging from 9 to 11 g/dl have approximately the same density of the brain on CT scans. What appears to be essential for an increase in CT attenuation is clot retraction with separation of serum and absorption of fluid, which in turn increase the hemoglobin concentration and thus the density of the clot.

The “hyperacute” hematomas are hyperdense, with some small areas of iso – or hypodensity within the lesion. The possible causes of these combined densities are the presence of fresh, unclotted blood (which has a low attenuation coefficient), a low hematocrit or a mix of blood with CSF due to tears in the arachnoid. Another proposed mechanism is the continuous washout of the blood within the hematoma through the diploic veins; however, the latter is usually associated with an overlying skull fracture and would apply to epidural hematomas and not subdural hematomas.

Low Attenuation Subdural Fluid Collections

Trauma in addition of giving rise to subdural hemorrhage can also cause an accumulation of nonhemic fluid. These have been identified in CT scans, which will show the nonhemic fluid collections as low attenuation subdural fluid collections as compared to clotted blood or the normal brain. The reason the CT scan demonstration of low attenuation subdural fluid is important in an infant or toddler is because it may represent a chronic subdural hematoma, suggesting remote trauma. In essence the identification of both acute subdural hematoma, or for that matter any intracranial hemorrhage in association with a low attenuation subdural fluid collection raises the possibility that more than one episode of significant head trauma has occurred. Although, some would contend that such a finding is virtually pathognomonic of repetitive child abuse, it must be born in mind that low attenuation subdural fluid collection can occur during the acute phase of head trauma, and, thus if accompanied by intracranial hemorrhage can be mistakenly diagnosed as an acute and chronic subdural hematoma indicative of repetitive child abuse. Thus, appropriate interpretation of neuroimaging is extremely important when attempting to give estimates as to when the trauma occurred. This is especially true when the history is uncertain. What is important to remember is that low attenuation subdural fluid collections can occur after a single traumatic event with no evidence of an associated intracranial hemorrhage. These low attenuation fluid collections in children under 3 years of age typically develop within 4 days of the traumatic event.

The location of the low attenuation fluid collection was in the frontal region in 98% with parietal involvement in 85%. The fluid collections were unilateral in 16% and bilateral in 84%. 89% of the bilateral fluid collections were symmetrical.

The most common associated hemorrhage was subdural in 87%, epidural in 6% and subarachnoid in 2%. 6% of patients had no visible intracranial hemorrhage. In 52% of those with intracranial hemorrhage, the hemorrhage and the low attenuated subdural fluid collection were in separate locations, e.g. epidural vs. subdural spaces, or posterior interhemispheric subdural space vs. the frontal convexity subdural space. In 8% the clotted blood and subdural fluid were mixed together and in 40% the subdural fluid and blood were adjacent to one another.

The concept that trauma-related low attenuation subdural fluid collections indicate several weeks have passed since the injury has developed from observations of adults with acute subdural hematomas that progressed to chronic subdural hematomas. A true chronic subdural hematoma shows a temporal evolution in attenuation values on CT. During the “subacute” phase of the injury, clot lysis causes the high attenuation of the hematoma to slowly decrease until it is isodense to brain after 2 to 3 weeks. Eventually, the attenuation values fall until they are equal to, or slightly higher than, those of clear cerebrospinal fluid. Oozing of blood or serum from the neovascular membrane of a subacute or chronic hematoma may alter the attenuation values.

Clot lysis cannot serve as an explanation for the formation of nonhemic subdural fluid. This is based upon the fact the nonhemic subdural fluid appears within 4 days of the traumatic event rather than weeks as is true of a chronic subdural hematoma.

Other possible pathogenic mechanisms for posttraumatic low attenuation subdural fluid accumulation include hygroma due to a tear in the arachnoid membrane, and effusion from traumatized meninges and a hyperacute hematoma with fresh blood or areas of unretracted semiliquid clot.

In one study, which involved 58 patients, 10 of which were children younger than 10 years with “Traumatic Subdural Hygroma” found the mean time at diagnosis was 11.6 days after injury. In another study utilizing MRI involving 2 children one “Traumatic Subdural Hygroma” was identified 3 days after the injury in a 7 week-old infant and 15 days after injury in a 4 year old. Another study involving 196 head injury patients, the hygromas were discovered at a mean of 22 days (range, 6-46 days).

It is also important that fluid within the subarachnoid space be differentiated from that in the subdural space, because the former is of little significance and the latter indicates a pathologic abnormality. The subarachnoid spaces in normal infants are frequently slightly prominent, and even greater prominence occurs with the self-limited entity of “benign extra-axial fluid collections of infancy.”

Infants who present with “benign extra-axial fluid collections” typically show macrocrania, mild ventriculomegaly and excessive extra-axial fluid. These infants have normal development and generally no treatment is necessary.

In one study involving 55 patients with “low attenuation subdural fluid collection,” 15% died within 30 days of the injury. 18% showed persistence of the fluid for at least 1 month. Spontaneous resolution occurred in 25%. Subdural shunts were placed in 15%.

Thus, most traumatic-related low attenuation subdural fluid collections in children younger than 3 years develop within the first week, typically 4 days, after the head injury. Coexistent clotted subdural blood is often present during the acute phase of the injury, and the imaging appearance may mimic that of acute hemorrhage into a chronic fluid collection. Therefore, although low attenuation subdural fluid collections in association with acute intracranial hemorrhage may represent multiple traumatic events, they are not diagnostic. The pathogenesis of these fluid collections is most likely multifactorial.