Subdural Hygromas in Abusive Head Trauma: Pathogenesis, Diagnosis, and Forensic Implications

SUMMARY: Are subdural hygromas the result of abusive head trauma? CT and MR imaging represent important tools for the diagnosis of abusive head trauma in living infants. In addition, in-depth understanding of the pathogenesis of subdural hygromas is increasingly required by neuroradiologists, pediatricians, and forensic physicians. Therefore, the current knowledge on subdural hygromas is summarized and forensic conclusions are drawn. The most important diagnostic pitfalls, benign enlargement of the subarachnoid space, and chronic subdural hematoma, are discussed in detail. Illustrative cases from forensic practice are presented. Literature analysis indicates that subdural hygromas can occur immediately or be delayed. If other infrequent reasons can be excluded, the presence of subdural hygromas strongly suggests a posttraumatic state and should prompt the physician to search for other signs of abuse. To differentiate subdural hygromas from other pathologies, additional MR imaging of the infant's head is indispensable after initial CT scan.

The diagnosis and dating of AHT in living infants predominantly relies on neuroimaging by means of CT and/or MR imaging. [23][24][25][26][27][28][29][30] An important indicator for AHT is the presence of SDHs attributed to defects of the bridging veins (BVs). Moreover, in some cases suspected for AHT, radiologists are confronted with almost homogeneous fluid collections within the subdural space, which appear isodense/isointense or nearly isodense/isointense to CSF. These collections are then interchangeably termed as subdural hygromas (SDHys), chronic subdural hematomas (cSDHs), or a terminologic mixture of both: chronic hygromas. This radiologic diagnosis has often resulted in difficulties with regard to the medicolegal assessment of such cases because the exact pathogenesis, diagnosis, and significance of SDHys are still a matter of debate and uncertainty, particularly in infants. 31,32 However, these issues can be decisive in court, for example, when SDHys are considered as evidence for the age of injury.
Therefore, the present review addresses the following questions: 1. What are SDHys? 2. What are the current theories regarding the pathogenesis of SDHy? 3. Which alternative explanations and differential diagnoses have to be considered? 4. What forensic implications arise concerning AHT?

What are Subdural Hygromas?
Pathology is always based on anatomy. Under physiologic conditions, the subdural space does not exist in humans. It does not open until the dura-arachnoid interface is mechanically separated, for instance due to brain shrinking, trauma, or neurosurgical interventions. [33][34][35] This opening is actually regarded as a cleaving of the so-called dural border cell layer-the innermost zone of the dura mater-and therefore also referred to as intradural lesion. 33,36 Nevertheless, in this article, the traditional term subdural is used to describe accordant pathologies within that space such as SDHy.
SDHys can be regarded and examined from 2 different points of view: first, the traditional perspective of the neuropathologist, forensic pathologist, or neurosurgeon who directly investigate or treat colorful 3D pathologies in the human head; and second, the perspective of the neuroradiologist who indirectly evaluates and interprets 2D black-and-white crosssectional images from CT, MR imaging, or sonography.
In radiology, the definition of SDHy is more difficult, and the terminology is very heterogeneous. While acute SDH, representing 1 of the leading indicators for Shaken Baby syndrome, can be reliably diagnosed by means of CT and MR imaging, other pathologic fluid collections are often termed interchangeably as SDHys, cSDHs, subdural effusions, chronic hygromas, or simply subdural collections.
The term subdural collection is used as an unspecific umbrella term for pathologically formed subdural fluid, whereas the term chronic hygromas should be principally avoided as it is a very imprecise and pathogenetically insufficient description. If there is a mass of proteinaceous liquid within the subdural space that appears to be associated with bacterial meningitis, it is generally spoken of as a subdural effusion. [54][55][56] But what about SDHy versus cSDH? A current neuroradiologic textbook by Osborn 55 defines SDHys as "hypodense, CSFlike, crescentic extraaxial collections that consists purely of CSF, have no blood products, lack encapsulating membranes, and show no enhancement following contrast administration" (Fig 1). This description is strikingly similar to the second morphologic type of SDHy suggested above by Unterharnscheidt. 37 However, this CSF-like appearance is also the reason why SDHys harbor a high potential to be confused with cSDHs. 57 According to Osborn, 55 cSDHs may be defined as "encapsulated collections of sanguineous or serosanguineous fluid confined within the subdural space." This description, on the other hand, closely resembles the first of Unterharnscheidt's 37 morphologic types of SDHy suggesting that cSDHs diagnosed by radiologists can also be termed as SDHys. In fact, the terms cSDH (in the meaning of old SDH) and SDHy are frequently used as synonyms in radiologic reports as well as in recent scientific literature. 34 Thereby, it is implied that SDHys solely indicate remnants of SDHs.
But is this simplification true? This may become an important issue for the forensic expert in court. Once the radiologic diagnosis of SDHy is made, the forensic expert will likely be confronted with 2 questions: Does the SDHy represent a result of AHT? And if yes, does it indicate recent injury, old injury, or, when in combination with other types of subdural collections, the presence of multiple injuries that occurred at different times? Therefore, it is important to understand how SDHys develop or what they originate from.

The Pathogenesis of Subdural Hygromas: An Ongoing Odyssey
Among neurotraumatologists it is generally known that SDHys usually derive from head injuries and represent rare posttraumatic complications that may coexist with epidural or subdural hematomas. 33,42,43,[57][58][59][60] Unfortunately, the causes of SDHys cannot always be read directly from the CT or MR images. Numerous scientists from different disciplines sought to address this problem in the last decades. Particularly in infants, SDHys are not well described and only little understood. Children were even excluded in a recent radiologic SDHy study because their pathogenetic aspects were considered a priori as different from adults. 61 Hereafter, the 2 current basic concepts of SDHy formation with their different medicolegal implications, as well as important alternative explanations, are presented.

Concept 1: Delayed Formation of Subdural Hygromas
It appears to be widely presumed that SDHys represent liquefied and/or deposited remnants of a previous acute SDH 52,62-65,82 suggesting that, in a case of suspected child abuse, the baby could have been abused weeks ago. In 1857, Rudolph Virchow considered SDHys as "final stages of subdural hemorrhages." 37 But are SDHys really direct remnants of acute SDHs?
An advanced approach was developed considering additional aspects of SDHy formation. The suggested process describes the conversion of acute SDH into cSDH via SDHy as an intermediate stage (blue pathway in Fig 2). 47,66 Because most acute SDHs re-solve rapidly, reflecting the high levels of tissue thromboplastin in brain tissue and CSF, 67 this approach has been refined by other authors. During the dissolving of the acute SDH, especially if decreased intracranial pressure is present, the cleaved dura-arachnoid interface is assumed to remain as persistent posttraumatic space. Liquid remnants of the acute SDH or CSF might then pass into that opened space by effusion from surrounding vessels or even the subarachnoid space, forming the SDHy. 50,52 It should be noticed that in this approach, the SDHy is considered as a consequence of the SDH but not as a directly transformed remnant of it.
Alternatively, Mack et al 36 suggested that CSF could physiologically move from the subarachnoid space into interstitial spaces of the dura mater and subsequently via the dural venous plexus into the dural sinuses. CSF might therefore be present in small amounts in the dura at all times. In any cases of alteration of this CSF absorption pathway-for instance from bleeding into the dural layers-a disruption of the transport mechanism may result in delayed accumulation of CSF within the subdural space producing imaging findings of SDHy. Approximately 30 years earlier, it was hypothesized that concurrent traumatic subarachnoid hemorrhages, that are frequently found together with SDHy, may secondarily predispose to defective CSF resorption leading to enlargement of subdural CSF collections as well. 43,68,69 In the end, SDHys represent a subsequent result of acute SDHs.
Following the concept of delayed SDHy formation, the pres- ence of SDHy may be medicolegally interpreted as later consequence of AHT that occurred some days or a few weeks before.

Concept 2: Rapid Formation of Subdural Hygromas
This theory suggests SDHy to be caused by a laceration in the arachnoid. 33,[37][38][39][49][50][51]53,[70][71][72] Hence, SDHy can result from trauma and from cranial operations in which the arachnoid membrane is opened, leading to CSF transfer into the subdural space. 33 For this, the term acute SDHy has been suggested. 52 A ball-valve mechanism of the arachnoid is assumed to prevent a backflow of CSF into the subarachnoid space. 33,72 In regard to a 74-year-old woman with acute SDHy and subarachnoid hemorrhage after accidental head injury, Kamezaki et al 73 discussed a tear of the arachnoid membrane as possible reason for the SDHy. Radiotracer and cisternography studies performed in infants with SDHy were able to show that radioisotopes (eg, indium-111) injected into the subarachnoid space move into the subdural compartment. 44,53 In addition, the CSF-specific ␤-trace protein and other proteins have been found up to 100% of SDHys, 74,75 indicating that there really is CSF in the subdural space.
Zouros et al 53 reported on 5 infants with suspected AHT showing acute hematohygromas. Encapsulating neomembranes, possibly indicating the presence of cSDH (see below: "Differential Diagnoses of Subdural Hygromas"), were not found either on MR imaging or at surgery. In addition, free CSF communication between subarachnoid and subdural space was proved in all cases by radiotracer injection. The authors proposed a mechanism whereby SDHy directly originates from shaking the baby: during acceleration and deceleration of the brain, acute shear strains be-tween arachnoid and dura may disrupt both the BVs and the weak arachnoid attachments to the parasagittal dura (Pacchioni granulations). Thereby, a mixture of CSF and blood products can flow into the traumatically opened subdural space resulting in an acute subdural hematohygroma. 53 This is not only a plausible explanation for the known mixed-density appearance of subdural collections frequently seen in AHT cases. The injured arachnoid granulations also explain the often-described occurrence of enlarged subarachnoid spaces because of suboptimal CSF absorption.
Following this concept of rapid formation, SDHys must not be considered automatically as direct remnants or delayed consequence of acute SDHs but SDHy and acute SDH may develop simultaneously as exemplified in Figure 3. Medicolegal expert opinions should therefore consider the possibility of an rapid formation of SDHy as an additional symptom indicating AHT.

Alternative Explanations
Besides the 2 main theories introduced above, additional theories regarding the formation of SDHys are discussed. In terms of forensic issues, 2 in particular are noteworthy: 1. Glutaric aciduria type I: this hereditary disease is caused by a deficiency of the enzyme glutaryl CoA-dehydrogenase and leads to increased urinary excretion of glutaric and 3-hydroxy-glutaric acid. 76 Clinically, macrocephalia and extrapyramidal movement disorders are described. In neuroradiology, frontotemporal atrophy as well as SDHys and/or SDHs are diagnosed. 77,78 Glutaric aciduria type I should therefore be diagnostically excluded in infants with SDHy, because misdiagnosing as AHT may occur in exceptional cases. 79,80 2. Rupture of pre-existing arachnoid cysts: arachnoid cysts are congenital or acquired intra-arachnoidal CSF collections occurring infrequently. These may rupture because of minor, but identifiable, head trauma, or sudden temporary rise in intracranial pressure. 48,81 Ruptures result in SDHys rather than in SDHs. 48 However, this phenomenon has not been described in infants yet. A review of the literature by Gelabert-González et al 81 demonstrated ages of occurrence ranging from 5 to 25 years. It seems conceivable that arachnoid cysts also rupture in infants and are not yet acknowledged as such. However, as long as scientific data do not support this possibility, this remains mere speculation.

Differential Diagnoses of Subdural Hygromas
Chronic Subdural Hematoma. Surgeons and pathologists know cSDH as subdural liquid with a dark brown "crank case oil" ap- pearance. 47 Many cSDHs also contain a mixture of both CSF and blood, such as breakdown products of hemoglobin or other proteins. 55,82 Furthermore, multiple hemorrhages of different ages are supposed to be common (so-called mixed-age SDH). 55 This may sometimes also lead to an attenuation approximating that of CSF. The pathogenesis of cSDH is not yet clear. cSDHs were observed to evolve directly from acute SDHs in only very few cases (reviewed in 52 ). Moreover, experimental studies failed to reproduce cSDH from acute subdural blood. 83,84 It has also been reported that, based on histopathology 85 and CT, 66 chronic and acute SDH should actually be regarded as different entities. cSDH might therefore not be (at least not directly) the last stage of an old acute SDH.
A direct transformation from remnants of an acute SDH into a cSDH is therefore not plausible in all cases. Instead, a more complex pathomechanism must be assumed. As mentioned above (Fig 2), cSDHs were commonly shown, at least in adult trauma cases (mostly traffic crashes), to derive from SDHys with incidence rates between 8% and 58%. 33,47,58,59 Thus, the fate of SDHy is either resolution or cSDH formation. 33,59,75 Under a normal pressure situation, the SDHy resolves. If the decreased intracranial pressure that had led to SDHy formation continues, the SDHy may expand. 50,52 This can enlarge the intradural cleavage (ϭ subdural space filled with SDHy) up to the opposite brain side opening a subdural space even above the falx cerebri. Accordingly, SDHys were frequently observed not to be restricted to the brain side of the "original pathology" (eg, acute SDH). 45 SDHy can then develop neomembranes from the proliferating dural border cells that are principally able to proliferate in any pathologic process with cleavage of the dural borderzone tissues. 50 Forming of neomembranes is accompanied by neovascularization. Spontaneous microhemorrhages from these fragile new vessels may then occur and lead to a mixture of CSF and blood. 86,87 Therefore, it has been suggested that repeated microhemorrhages possibly convert an SDHy into an expanding cSDH. 33,50,52,57,66 These pathogenetic considerations show why it is important for the forensic expert to differentiate between cSDH and SDHy. While cSDHs appear to be very rare and delayed consequences of subdural collections, SDHys can apparently develop delayed or rapidly. However, differentiation can be impossible for the radiologist in cases of cSDH appearing CSF-like in CT or MR imaging. These cSDHs are particularly vulnerable to be accidentally referred to as SDHys. Consequently, the terms cSDH and SDHy are often used as synonyms in practice.
Some authors describe the differences between cSDH and SDHy as follows: SDHys are thought to be less than 3 weeks old, static or decreasing, and do not or rarely produce a mass effect, whereas cSDHs are thought to be older than 3 weeks, enlarging, and may cause a mass effect. 50,51,88 However, all these smooth differencing criteria should be handled with care. "Three weeks" cannot be a strict borderline, and most of the aforementioned pathophysiologic data regarding cSDH rely on studies in adults. Thus, it remains at least questionable whether these results can be applied to infants at all.
As SDHys mostly lack neomembranes, this aspect could be another morphologic criterion for differentiation. Neomembranes are usually present in cSDHs and encapsulate the subdural collection as a result of tissue response and may even subdivide it into different chambers. 34,50,55 Although neomembranes are described as becoming visible to the naked eye after approximately 10 days 34 and were shown to aid in dating injuries, 89 diagnosing such membranes in CT or MR imaging can be very challenging.
Benign Enlargement of the Subarachnoid Space. The benign enlargement of the subarachnoid space (BESS) represents an important differential diagnosis for both SDHy and cSDH ( Fig  4). 48,51,52,90 These subarachnoid fluid collections are frequently observed and often termed confusingly as "benign hygromas of infancy." BESS probably results from immaturity of the arachnoid villi leading to a transient form of communicating or external hydrocephalus. 52 Infants concerned are usually neurologically uneventful without evidence of prior brain injury. 91 Nowadays, BESS can clearly be distinguished from SDHy (Fig 5), particularly because of improvements in MR imaging technology. In the presence of BESS, the vessels, which run through the subarachnoid space, are localized away from the brain. On the other hand, in the presence of a subdural fluid collection, the vessels can be found near the surface of the brain. Forensically, it is important to know that long-term observations of infants with BESS as well as a finite element study indicated no increased risk for developing SDH. [92][93][94][95] The hypothesis was that stretching of the BV due to enlargement of the subarachnoid space may result in a predisposition to developing SDH. Accordingly, it has been reported that SDHs may occur either spontaneously or as a result of minor trauma in infants with BESS. 96 By contrast, it has been frequently discussed that enlarging of the head circumference because of external hydrocephalus is rather a consequence and not the cause of bleeding, for instance, due to impaired CSF absorption caused by subarachnoid pus, cells, or hemorrhage, or by SDH. 51,54,95 Besides BESS, it is essential to know that, in general, the subarachnoid spaces are relatively larger in the first 2 years of life than in older children or adults. 52,82,87 According to Libicher and Tröger, 97 head sonography of 89 healthy American infants revealed the distances between the inner calvarian table and the cerebral cortex to range from 0.3 to 6.3 mm (upper limit proposed based on the 95th percentile: 4 mm).The infants' head circumferences of that study were found to be between the 3rd and 97th percentile. In addition, cerebral atrophy, for example, as a result of AHT, may also lead to the impression of enlarged subarachnoid spaces. 52,82 CONCLUSIONS SDHy and cSDH are often difficult to distinguish from each other and are often used synonymously in daily case work. While cSDHs in infants are rare and rather implicate a delayed and nonacute process, SDHys may develop rapidly or be delayed. Accordingly, early hypodensity in infantile SDH has also been observed by others 51,98-100 arguing against an overhasty diagnosis of a chronic process but suggesting a significant role of CSF.
On one hand, SDHys are classically considered as remains of a previous SDH, directly or indirectly, which strongly suggests a trauma of cortical BVs and a delayed formation of SDHys (see Concept 1 section). On the other hand, acute pathogenesis of SDHys has been verified by traumatically induced tears in the arachnoid membrane (see Concept 2 section). In all probability, multiple mechanisms exist and also coexist. Accordingly, it is not adequate to state different ages of injuries when SDH and SDHy are present concomitantly.
Both concepts presented have 1 thing in common: if other infrequent reasons and differential diagnoses have been excluded, the presence of SDHy strongly suggests trauma, or more precisely: a posttraumatic state. The presence of SDHy in infants therefore represents compelling reason to search for other signs of AHT such as retinal hemorrhages, fractures, bruises, or inadequate explanations for trauma.
The usage of SDHy for age estimation of head trauma is difficult and should not be considered as the most important factor in determining the time of injury. 53 Hence, as already proposed by Vezina,51 in initial CT investigations, it is best to describe subdural collections only in terms of density (hypo-, hyper-, isodense, or mixed) and strongly avoid labels such as "acute" or "chronic." If additional presurgical MR imaging scans of the head exist, further assessment is possible. MR imaging is clearly more sensitive to the presence of SDH, BESS, neomembranes, and injuries of the cerebrum, brain stem, or upper cervical cord. Furthermore, temporal development of intracranial hemorrhages by means of MR imaging is well studied. 51,52 To conclude, evaluation of SDHy cases should ideally be done in close cooperation between neuroradiology, pediatrics, and forensic medicine. The initial neuroradiologic evaluation is of particular importance and indispensable for correct medicolegal conclusions.