From the Harvard Medical School and Children's Hospital, Boston, Mass.
Brain injury in the premature infant is an extremely important problem, in part because of the large absolute number of infants affected yearly. The 2 principal brain lesions that underlie the neurological manifestations subsequently observed in premature infants are periventricular hemorrhagic infarction and periventricular leukomalacia. The emphases of this article are the neuropathological features, pathogenesis, and potential means of prevention of these 2 lesions. Recent work suggests that the ultimate goal, prevention of the lesions, is potentially achievable. Periventricular hemorrhagic infarction may be avoidable by prevention of germinal matrix–intraventricular hemorrhage, and periventricular leukomalacia by detection of impaired cerebrovascular autoregulation, prevention of impaired cerebral blood flow, and interruption of the cascade to oligodendroglial cell death by such agents as free radical scavengers.
The neurologic outcome of infants born prematurely represents a problem of enormous importance. Approximately 50000 infants are born yearly in the United States with a birth weight of less than 1500 g. This group represents approximately 1.2% of live births, and this proportion has not changed appreciably in the past 10 years in the United States. Because of major advances in neonatal intensive care, approximately 85% of these small infants survive. However, of the survivors, approximately 5% to 15% later exhibit spastic motor deficits, categorized as cerebral palsy.1 Indeed, the prevalence of cerebral palsy has increased in many countries in recent years in considerable part because of the contribution of the survivors of premature birth.2 Additionally and importantly, 25% to 50% of survivors also exhibit abnormalities of cognition and behavior, with learning disturbance the nearly uniform result.3 The emphasis of this article is on the neuropathological substrate of these common neurologic abnormalities, our current concepts of pathogenesis, and the potential means of prevention of the injury.
The principal neuropathological substrates for the neurologic disturbances in premature infants involve the cerebral white matter. These lesions, periventricular hemorrhagic infarction and periventricular leukomalacia, are the focus of this article. However, contributory neuropathologic conditions include selective neuronal injury, occurring particularly in a pattern affecting pontine tegmentum and subiculum (pontosubicular necrosis), posthemorrhagic hydrocephalus, and abnormalities of the subsequent development of cerebral cortical circuitry after neonatal white matter injury.1,4
The neuropathological characteristics of periventricular hemorrhagic infarction are striking and consist of a relatively large region of hemorrhagic necrosis in the periventricular white matter, just dorsal and lateral to the external angle of the lateral ventricle.1 The necrosis is strikingly asymmetrical: in the largest series reported, 67% of such lesions were exclusively unilateral and in virtually all the remaining cases, grossly asymmetrical, although bilateral.5 Approximately 80% of cases are associated with large germinal matrix–intraventricular hemorrhage (GMH-IVH), and commonly (and mistakenly) the parenchymal hemorrhagic lesion is described as an "extension" of the GMH-IVH. That simple extension of blood into cerebral white matter from germinal matrix or lateral ventricle does not account for the periventricular hemorrhagic necrosis has been shown clearly by several neuropathological studies.3- 5 Microscopic study of the periventricular hemorrhagic necrosis indicates that the lesion is a hemorrhagic infarction.1 Moreover, the hemorrhagic infarction tends to be most concentrated near the ventricular angle where the medullary veins draining the cerebral white matter become confluent and ultimately join the terminal vein in the subependymal region. It is now apparent that periventricular hemorrhagic necrosis occurring in association with large GMH-IVH is, in fact, a venous infarction. This lesion is distinguishable neuropathologically from secondary hemorrhage into periventricular leukomalacia, the ischemic, usually nonhemorrhagic, and symmetrical lesion of periventricular white matter of the premature infant (see next paragraph). However, distinction of these 2 lesions in vivo often is difficult, and, indeed, they frequently overlap in the same brain. A comparison of the basic features of these 2 periventricular white matter lesions of the premature infant is as follows: periventricular hemorrhagic infarction is nearly invariably unilateral or markedly asymmetrical, invariably grossly hemorrhagic, and has a probable venous site of circulatory disturbance; in contrast, periventricular leukomalacia is rarely unilateral or markedly asymmetrical, rarely grossly hemorrhagic, and has a probable arterial site of circulatory disturbance.
The pathological features of periventricular leukomalacia are distinctive and consist primarily of both focal necrosis and more diffuse cerebral white matter injury.1 The focal necrotic lesions deep in the cerebral white matter are characterized by coagulation necrosis with loss of all cellular elements, ie, an infarction. The earliest feature of the focal necrotic lesions is the appearance of round neuroaxonal swellings (retraction clubs and balls). Thus, there is evidence for axonal rupture and, potentially, extravasation of substantial amounts of glutamate, present in millimolar concentrations in neurons, into the periventricular white matter (see the following section on pathophysiology). The more diffuse cerebral white matter injury is characterized by loss of oligodendrocytes, specifically, with a corresponding increase in hypertrophic astrocytes. The cellular targets appear to be early differentiating oligodendrocytes. Subsequent loss of these cells and resulting impairment in myelination, manifested by a diminished volume of cerebral white matter and an increase in ventricular size, are the later correlates.
The pathogenesis of periventricular hemorrhagic infarction appears in most cases to be related directly to the associated GMH-IVH on the basis of several key, previously reported observations.5 First, approximately 80% of the parenchymal lesions were observed in association with large (and usually asymmetrical) GMH-IVH. Second, the parenchymal lesion invariably occurred on the same side as the larger amount of germinal matrix and intraventricular blood. Third, the parenchymal lesion developed and progressed after the occurrence of the GMH-IVH. These data raised the possibility that the GMH-IVH led to obstruction of the terminal veins and thereby hemorrhagic venous infarction. This pathogenic formulation recently received strong support from Doppler ultrasonographic determinations of blood flow velocity in the terminal vein during the evolution of the infarction in the living premature infant; obstruction of flow in the terminal vein by the ipsilateral GMH-IVH was shown clearly.6
Concerning prevention of periventricular hemorrhagic infarction, this discussion of pathogenesis leads to the conclusion that the major cause is GMH-IVH. Thus, prevention of the infarction centers around prevention of the initiating GMH-IVH. Although this topic is discussed in detail elsewhere,1 most data support a beneficial role in prevention of GMH-IVH for the use of antenatal corticosteroids, muscle relaxation therapy for the infant breathing out of synchrony with a ventilator, and careful postnatal management of ventilation and perfusion. Encouraging evidence is available for the postnatal use of indomethacin, vitamin E, or ethamsylate, but at present the data are not sufficiently conclusive to recommend routine prophylactic use of these agents.
The pathogenesis of periventricular leukomalacia is related to 3 major interacting factors that result in ischemia to the periventricular region and injury to the particularly vulnerable cerebral white matter. The first set of pathogenic factors of importance is the periventricular vascular anatomical situation in the premature infant. The occurrence of periventricular leukomalacia in areas that represent maturation-dependent arterial end zones or border zones of the major penetrating arteries has been emphasized by many authors.7- 9 These arterial end and border zones are essentially "distal fields," ie, watershed areas. As such, these zones would be expected to be most susceptible to a decrease in perfusion pressure and cerebral blood flow, as discussed elsewhere.3 The arterial end zones in deep periventricular white matter are considered to be the site of severe ischemia and thereby infarction with necrosis of all cellular elements. Less pronounced arterial border and end zones also appear to exist in the cerebral white matter relatively distant from the periventricular region and thereby, with decreases in cerebral blood flow, subject the cerebral white matter to a moderate ischemic insult and injury specifically to oligodendrocytes.
THE SECOND major pathogenic factor relates to the propensity for cerebral ischemia and a pressure-passive cerebral circulation to occur in the premature infant. In the presence of impaired cerebral blood flow the vascular end zones and border zones just described thus render the brain particularly vulnerable to injury. Perhaps of greatest importance in the genesis of impaired cerebral blood flow and thereby cerebral ischemia is an impairment of cerebrovascular autoregulation in certain premature infants.10 The pressure-passive abnormality of the cerebral circulation in such premature infants relates in part to an absent muscularis around penetrating cerebral arteries and arterioles in the third trimester in the human brain.
The difficulty in conclusively establishing a relationship between a pressure-passive cerebral circulation and the occurrence of periventricular white matter injury in the premature infant has related to the inability to measure quantitative changes in the cerebral circulation from second to second. The advent of near-infrared spectroscopy has changed this situation.3 Thus, this noninvasive technique allows the measurement, essentially in real time, of cerebral concentrations of oxygenated and deoxygenated hemoglobin. Changes in the concentrations of these 2 intravascular compounds provide important information about the cerebral circulation. Using this technique in a preliminary study of 35 premature infants from the first hours of life, my colleagues and I have identified a subset of infants (26% of the total group) with a pressure-passive cerebral circulation and an extremely high incidence (100%) of the development of periventricular leukomalacia (or IVH). These observations suggest that premature infants with a pressure-passive cerebral circulation are at high risk for the development of ischemic white matter injury and that such infants can be identified prior to the occurrence of the injury. Current work is directed at identification of the causes of the cerebral circulatory abnormality and the means of prevention of this disturbance.
However, it should be recognized that even in the presence of an intact cerebrovascular autoregulation, marked cerebral vasoconstriction potentially could lead to impaired cerebral blood flow to the periventricular vascular end zones and border zones. This explanation may account for the demonstrated relationship between marked hypocapnia and periventricular leukomalacia in premature infants.11
The third important factor in the pathogenesis of periventricular white matter injury relates to an intrinsic vulnerability of neonatal cerebral white matter. Such an intrinsic vulnerability of the immature oligodendrocyte in cerebral white matter of the human infant, particularly regarding the more diffuse component of periventricular white matter injury, is suggested by experimental studies, by the rarity of the lesion at later ages, and by the relative cellular specificity of the diffuse injury, ie, involving oligodendrocytes but not astrocytes or other cellular elements.1,12 Recent immunocytochemical studies of developing human brain suggest that the cellular target for diffuse oligodendroglial injury in periventricular leukomalacia is an early differentiating oligodendrocyte, ie, at a developmental stage prior to the acquisition of myelin basic protein staining typical of the mature oligodendrocyte.12 Studies in cell culture demonstrate that this early differentiating oligodendrocyte, but not the mature oligodendrocyte, is extremely vulnerable to free radical attack.13,14 This vulnerability, of course, is of great interest because periventricular leukomalacia is considered to be an ischemic lesion, and an elevation in a variety of reactive oxygen species is a well-established sequela of ischemia and/or reperfusion. In 2 model systems of free radical accumulation, my colleagues and I have shown that early differentiating oligodendrocytes are very vulnerable to free radical attack.13,14 In one model, oligodendrocytes were shown to undergo free radical–mediated cell death by exposure to glutamate.13 The mechanism of the glutamate-induced cell death was related to glutathione depletion caused by intracellular cystine depletion, in turn related to activation of a glutamate-cystine exchange transport system. Glutamate entry caused cystine efflux, glutathione depletion, and free radical–mediated cell death. Clinically safe free radical scavengers, eg, vitamin E, totally prevented the oligodendroglial death caused by glutamate.
The mechanism of the intrinsic vulnerability of early differentiating oligodendrocytes to free radical attack remains to be defined but could relate to a delay in development of antioxidant defenses, especially catalase and glutathione peroxidase, during an interval of rapid acquisition of iron that is crucial for oligodendroglial differentiation.12 Indeed, we have obtained support for this mechanism in early differentiating oligodendrocytes by the demonstration that the free radical–mediated oligodendroglial death can be completely prevented by the presence of an iron chelator, deferoxamine mesylate.14 This finding is highly relevant to the human brain in vivo, because iron and transferrin are crucial for oligodendroglial differentiation and for subsequent myelination.12,15 The oligodendrocyte has been shown to be the brain cell type most enriched in iron and the iron-binding proteins transferrin and ferritin. Moreover, the presence of ferritin-positive oligodendrocytes in cerebral white matter has been demonstrated as early as 25 weeks of gestation.16 Taken together, the data suggest a maturation-dependent window of vulnerability to free radical attack during oligodendroglial development because of the active acquisition of iron for differentiation at a time of relative delay in the development of certain key antioxidant defenses.
Other potential mechanisms for oligodendroglial injury also may contribute to the cell's vulnerability. For example, Yoshioka and coworkers17 have shown that oligodendrocytes express α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors and that cell death can be produced by activation of such receptors. This observation suggests a second potential mechanism of oligodendroglial cell death produced by glutamate (see earlier discussion).
A potential role for cytokines in the genesis of periventricular white matter injury has been suggested because of the relationship of the lesion to maternal infection and chorioamnionitis.12 Moreover, certain cytokines, eg, tumor necrosis factor α and interferon gamma, have been shown to be toxic to oligodendrocytes, although the data are not entirely consistent for tumor necrosis factor α.12
Concerning potential preventive interventions for periventricular leukomalacia, especially critical is maintenance of cerebral perfusion. In the infant with intact cerebrovascular autoregulation, prevention of severe hypotension and avoidance of marked cerebral vasoconstriction, eg, marked hypocapnia, are important. In the infant with a pressure-passive cerebral circulation, early detection of the cerebrovascular abnormality, eg, by near-infrared spectroscopy, is a vital starting point. Correction of the pressure-passive state requires recognition of cause, eg, marked hypercapnia or hypoxemia, infant breathing out of synchrony with the ventilator, etc, and then appropriate intervention. Perhaps of greatest importance will be interventions to prevent the cascade to oligodendroglial death related to free radical attack. Such agents as free radical scavengers may ultimately prove to be critical. Initial enthusiasm for antenatal administration of magnesium,18,19 a simple compound with antioxidant (as well as antiexcitotoxic and vasoactive) properties, has been tempered somewhat by the negative results of a recent, large study of the effect of maternal receipt of magnesium on neonatal white matter injury.20 Additional data are needed.
Accepted for publication July 24, 1997.
Reprints: Joseph J. Volpe, MD, Department of Neurology, Fegan 1103, Children's Hospital, 300 Longwood Ave, Boston, MA 02115 (e-mail: email@example.com).
Thank you for submitting a comment on this article. It will be reviewed by JAMA Neurology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Register and get free email Table of Contents alerts, saved searches, PowerPoint downloads, CME quizzes, and more
Subscribe for full-text access to content from 1998 forward and a host of useful features
Activate your current subscription (AMA members and current subscribers)
Purchase Online Access to this article for 24 hours
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 77
Customize your page view by dragging & repositioning the boxes below.
Users' Guides to the Medical Literature
Table 9.2-2 Refuted Evidence From Studies of Physiologic or Surrogate Endpoints
All results at
and access these and other features:
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.