动物造模,儿童虐待性脑外伤动物模型 z-bio 2022-10-19 284

　　Introduction: Abusive head injury (AHT) is currently recognized as the name of intentional head injury to young children, because the head injury is not attributed to any specific pathogenesis. This neurological disease is also known as "non accidental head injury in children", "swaying baby syndrome", "abused baby syndrome", "whipping swaying injury" and "swaying shock syndrome". Neuropathological changes of AHT include acute and chronic subdural hemorrhage (SDH), brain swelling, cerebral hemorrhage, contusion tear, hypoxic-ischemic injury, and traumatic and ischemic axonal injury (AI). The pathophysiology and biomechanics that produce these injuries individually or in various combinations vary from case to case. The acute encephalopathy pattern found in AHT is global and ischemic. Neuron damage is widely distributed, characterized by cytoplasmic atrophy and eosinophilia. Diffuse brain swelling caused by hypoxia is a common cause of death. Diffuse axonal injury (DAI) is uncommon in this age group, but vascular AI caused by elevated intracranial pressure is a common finding. There are two modes of ischemic axon injury: ischemic and traumatic. In infants less than 3 months old, the craniocervical (cervical spinal) junction seems to be particularly vulnerable to traumatic injury. It is found in AHT that ischemia and hypoxia in this area are considered the structural basis of children with apnea and/or dyspnea. This ischemic axonal injury may lead to cardiopulmonary arrest, followed by global cerebral ischemia and swelling. However, the explanation of this mechanism for unilateral brain injury is not very convincing. SDH in AHT cases is characterized by a thin layer of blood, usually distributed bilaterally. Shear force generated by head acceleration/deceleration is the most likely cause of bridge vein rupture of SDH, although hypoxic endothelial damage and increased vascular permeability in these immature veins are also considered as a hypothetical damage mechanism. Retinal hemorrhage (RH) in AHT is more common and severe than accidental injury in infants, but it is still non-specific because there are many other potential causes. AHT hemorrhage is usually extensive and involves all retinal layers. They are believed to be caused by rotational acceleration/deceleration forces, although some speculate that RH may be caused by hypoxic vascular injury.

　　AHT animal model: published AHT studies are often criticized for their case selection bias. The conclusion of systematic evaluation is that there is not enough scientific and evidence-based research to evaluate the diagnostic accuracy of "triad" in actively diagnosing AHT, and there is a dispute about whether triad can be independently shaken. The diagnosis of AHT is often hampered by the lack of a convincing historical record of the circumstances surrounding suspected cases of AHT due to confusion among the alleged perpetrators and the absence of reliable witnesses. The absence of contact damage does not necessarily mean that there is no head impact. One of the main objectives of animal modeling should be to better understand the pathogenesis and biomechanics of pathological changes found in AHT. When developing AHT models, many characteristics of children's brains must be considered. These attributes explain why it is often problematic to extrapolate the data of adult traumatic brain injury (TBI) studies to children. Incomplete myelination of infant brain may be the reason why the response to trauma is different from that of adults, especially the low shear value of myelin and other lipids. Increased susceptibility to shear injury due to incomplete myelination and immature astrocytes seems to be beneficial to diffuse brain injury. The skull vault of infants is soft and flexible, the sutures are not fused, the fontanel is open, the bone and dura mater are rich in blood vessels, the dura mater is not firmly connected with the internal surface of the skull, the relatively large subarachnoid cavity is rich in blood vessels, and the smooth bone support at the bottom of the skull has almost no resistance to brain movement. The higher the water content of immature brain, the lower its compressibility, but it is also more vulnerable to the impact of increased intracranial pressure, although this may be offset to some extent by the separation of immature fontanel and suture. The brain of an infant can withstand greater deformation than that of an older child, and usually requires greater force to deform the child's brain. The suture in the infant skull can deform more than 100% before rupture, but the resistance of the adult skull to mechanical force is 11 times higher because the stiffness of the skull increases with age. Although the more deformable skull may reduce the diffuse brain damage caused by angular acceleration, it may increase the local damage below the impact site. When extrapolating the results from one species to another, we must also consider the time of sudden growth, that is, the short period of rapid brain growth. It is important to compare the brains of different species at comparable developmental stages, not relative to the age at birth. According to the sudden increase of brain growth, mammalian species can be mainly divided into prenatal (e.g. sheep), perinatal (humans, pigs) and postpartum (rats, mice) brain development. There is a big gap in our understanding of the pathophysiology and biomechanical mechanisms involved in the development of AHT brain injury, and these studies cannot conduct ethical research on humans, but few animal models are designed to study the pathogenesis of AHT. The main focus of this review is to evaluate the utility of animal models that have been developed to study neuropathological changes in AHT. The evaluation of AHT biomechanical models is beyond the scope of this review, but they are very important for analyzing the head movement after injury, or the differential movement in the brain or between the brain and the skull. The research of shaking and head impact requires very different biomechanical modeling. The former has relatively low strength and long duration, while the latter has high strength and short duration. Due to different model types and injury mechanisms, it is often difficult to compare the results of different biomechanical models, and the experimental results are sometimes contradictory. Similar to biomechanical studies, it is difficult to compare the neuropathological changes found between different AHT animal models, because the range of mechanical devices developed is to produce head movements in different directions in different animal species. The head is usually confined to these devices to enhance the repeatability of results rather than free movement, and various histological techniques have been used to detect brain damage. This review attempts to assess the reliability of these models in reproducing the neuropathological changes associated with human AHT, including their occurrence consistency, severity, and extensive neuroanatomical distribution. The evaluated lesions were ischemic hypoxic neuronal damage, multifocal AI in different white matter tracts, blood brain barrier (BBB) destruction and diffuse angiogenic edema, as well as SDH/SAH and RH. If a lesion is not described in a given study, it is assumed that it has not been evaluated.

　　From this analysis, it is clear that the animal models developed so far have not reliably reproduced all the neuropathological changes found in abused human infants, although some models may be useful for studying the pathogenesis of specific lesions. The only reliable model that can produce extensive neuronal ischemia and hypoxia injury is the model developed by the University of Pennsylvania. The 3-5 day old piglets keep their heads rotating in different directions. In AHT lamb model, APP immunoreactivity was found in neurons widely distributed in brain and retinal ganglion cells. This neuronal response is considered to be a nonspecific acute stress response to manual shaking. Progressive neocortical neuronal degeneration was also found in shaken infant mice using silver immersion technique. Focal AI in different white matter tracts was detected in 8 to 12 day old shaking mice, 3-5 day old mechanical head rotating piglets, and 7 to 10 day old manually shaking lambs. AI in lambs is particularly severe in the white matter of the hemisphere, but it is also widely distributed at the junction between the brain stem and the skull neck (the place with the greatest load during the vibration). It is speculated that AI at the craniocervical junction may cause apnea and cardiopulmonary arrest in some dead lambs. The parenchymal damage of the mouse model was limited to the white matter around the ventricles, leading to hemorrhage and ultimately leading to cystic changes, although AI was more extensive. SDH/SAH is found in some animal models, but RH is uncommon. Three week old piglets injected exogenous blood into the subdural cavity through craniotomy, instead of generating SDH through mechanically induced head movement. The early gene c-fos was significantly expressed in the meningeal epithelial cells of the craniocervical spinal cord of the manually shaken lamb model, but also in the vascular endothelial cells of the hemisphere, cerebellum and brainstem meninges, meninges and white matter of the hemisphere. The pig model of the University of Pennsylvania was used to study the brain injury caused by mechanical head shaking in different directions and different application frequencies. Piglets were chosen because they are similar to human postnatal brain development sequence, have similar cerebrovascular development, and have similar responses to cerebral blood flow and EEG activity. From these pig studies, it was concluded that the severity of neuropathological changes, especially AI, depends on the number of sustained traumatic vibration injuries, the time interval between these injuries, and the direction of rotation of the head movement caused by the impact force. Traumatic AI of 3-5 day old piglets after repeated head rotation is greater than that after single head rotation, which means the cumulative effect of this traumatic injury. AI is also greater when the rotation interval is 24 hours instead of 7 days. Periodic low speed head rotation of piglets produces more AI than single rotation of the same amplitude. Piglets were also used in the same mechanical device to study brain injury caused by head rotation in different directions. Sagittal and horizontal (but not coronal) head movements often produce hypoxic-ischemic neuronal damage, multifocal AI, and SDH. Three quarters of the animals developed RH 6 hours after injury after head movements in different directions, and 70% of the bleeding was located in the vitreoretinal region. Axial rotation produced more hemorrhage than sagittal rotation or coronal rotation. These findings support the view that acceleration/deceleration force leads to abnormal traction of the vitreous body on the retina, thus damaging retinal blood vessels and causing hemorrhage.

　　Discussion: Up to now, the experimental animal models developed by laboratory rodents and domestic animals cannot reliably replicate all the neuropathological changes found in human infant AHT. This is largely due to the irreconcilable differences in neuroanatomical species and the fact that in mechanical devices that cause vibration damage in different directions, the head is usually restricted rather than free to move. Unfortunately, one of the common attributes of TBI animal models is that the more controlled and repeatable the mechanical input, the less likely the model is to be a real world human neural trauma, and the more problematic the experimental results are for human transformation. In most laboratory rodent models and pig models at the University of Pennsylvania, the head movement is limited to specific directions, which cannot simulate the head movement that may occur when a human baby is maliciously shaken. One of the most controversial issues in AHT that has not been resolved in animal models is whether shaking alone is enough to cause severe brain damage, or whether additional head impact is required.

　　Conclusion: It is impossible for any animal model to accurately reproduce the complete range of brain and eye lesions used to support the diagnosis of human infant AHT, and the selected model may be the most suitable model for the specific lesions under study. The huge difference between the immature experimental rodent brain and the human infant brain will often cause problems in the transformation of experimental results between species, especially when the animal head is restrained rather than moving freely in the process of shaking. Despite these difficulties, animal models have greatly improved our understanding of the pathogenesis of adult TBI, and are expected to be further improved to produce more reliable results of AHT neuropathology diagnosis.