Preterm Intraventricular Hemorrhage/Posthemorrhagic Hydrocephalus

IVH as an Accompaniment of Periventricular Leukomalacia

Periventricular leukomalacia, the generally symmetrical, nonhemorrhagic, and apparently ischemic white matter injury of the premature infant (see Chapter 18, Chapter 19, Chapter 20 ), was observed to some degree in 75% of one series of infants who died with IVH. The frequent association of classic necrotic/cystic periventricular leukomalacia and IVH also was emphasized in three other neuropathological reports, as well as in two large ultrasonographic studies. Although it has been reported that approximately 25% of examples of periventricular leukomalacia become hemorrhagic, especially when associated coagulopathy is present, this figure includes examples that have been accompanied by large IVH and that probably represent the venous infarction discussed earlier as periventricular hemorrhagic infarction. Takashima and co-workers suggested that the two lesions (i.e., periventricular hemorrhagic infarction and hemorrhagic periventricular leukomalacia) may be distinguishable in part on the basis of topography. Thus hemorrhagic periventricular leukomalacia has a predilection for periventricular arterial border zones, particularly in the region near the trigone of the lateral ventricles. Venous infarction, especially its most hemorrhagic component, is particularly prominent more anteriorly; that is, the lesion radiates from the periventricular region at the site of confluence of the medullary and terminal veins and assumes a roughly triangular, fan-shaped appearance in periventricular white matter.

IVH and Causation of Periventricular Leukomalacia/Cerebral White Matter Injury

IVH also may contribute to the occurrence of periventricular leukomalacia/cerebral white matter injury This possibility is supported by experimental, neuropathological, and clinical imaging studies ( Fig. 28.10 ).

An excellent series of experimental studies demonstrated deleterious effects of intraventricular blood and blood components on developing brain. The injury scenario begins with red blood cell lysis and hemoglobin release into the ventricular system. The ependymal injury related to the IVH likely facilitates the penetration of hemoglobin and heme into brain parenchyma. Heme released from hemoglobin is taken up especially by microglial cells and then degraded by heme oxygenase to iron, carbon monoxide, and biliverdin. Iron facilitates the formation of injurious reactive oxygen species. The latter are injurious to both pre-OLs and axons. Heme can also be taken up by contiguous neurons, which are then injured. Such neurons could include those adjacent to blood in the ventricles (e.g., hippocampus, subventricular zone, thalamus) or in the subarachnoid space (e.g., external granule cells of the cerebellum). Neurons are especially vulnerable because they do not contain ferritin, which can help sequester iron. Hippocampal neurons are especially affected. (Heme is involved in the subsequent development of hydrocephalus, although the mechanisms are not entirely clear.) The particular role of free iron in the scenario leading to injury is supported by the beneficial effects of systemically administered deferoxamine in an animal model of IVH. The latter agent led to reduced hippocampal neuronal loss and also to reduced occurrence of hydrocephalus.

Further insights have been provided by a novel animal model of IVH produced by intraperitoneal glycerol in administration of a preterm rabbit model to investigate the effect of IVH on cortical development. In this model, IVH pups had globally reduced myelin content, an aberrant cortical myelination microstructure, and thinner upper cortical layers (I–III). The authors also observed a lower number of interneurons in deeper cortical layers (IV–VI) in IVH animals and reduced numbers of neurons, synapses, and microglia. Dohare et al also demonstrated a significant reduction in neurons in the upper cortical layer (II–IV) in preterm IVH pups.

Additional blood products are important in leading to parenchymal injury after IVH. Thus after vessel rupture, prothrombin and fibrinogen enter brain and activate the coagulation cascade. The resulting thrombin promotes neuroinflammation via activation of microglia and astrocytes and thereby may lead to cellular injury. Thrombin specifically suppresses differentiation of pre-OLs into myelin-producing mature OLs. Fibrin also leads to microglial activation. These neuroinflammatory responses would be expected to injure vulnerable pre-OLs, axons, and contiguous neurons.

A final mechanism involves hyaluronan , generated from reactive astrocytes and damaged extracellular matrix. IVH leads to an increase in hyaluronan receptors, which activate microglia and directly inhibit pre-OL maturation. Hyaluronidase, injected intraventricularly, degraded hyaluronan and repressed the neuroinflammation, promoted pre-OL maturation, and restored myelination and neurological function in a preterm rabbit model of IVH.

Studies of IVH in human premature brain postmortem and in vivo , albeit relatively few, appear consistent with the experimental data. Elegant studies of the postmortem human brain have shown (1) in the subventricular zone of the ganglionic eminence, impairment of proliferation of oligodendroglial precursor cells and of neuronal cells, and (2) in the cerebral white matter, impairment of pre-OL maturation, (3) axonal injury, and (4) microglial activation. The pre-OL disturbance would be expected to lead to impaired myelination. The neuronal disturbance in the ventral ganglionic eminence likely involves late proliferating GABAergic neurons destined especially for thalamus. Notably, two MRI studies of relatively mild GMH-IVH in premature infants have shown subsequently impaired cortical and deep nuclear development. Whether the cortical disturbance could represent a dysmaturational effect caused by an impairment of axonal ensheathment and by direct axonal injury, as in nonhemorrhagic cerebral white matter injury (see Chapter 18, Chapter 19, Chapter 20 ), is unclear.

Concerning pathophysiology, as noted earlier, large amounts of non-protein-bound iron are present in CSF of infants after large GMH-IVH (with Post hemorrhagic hydrocephalus [PHHC]), and quantitative susceptibility map analysis by MRI with severe GMH-IVH shows changes consistent with accumulation of hemosiderin/ferritin iron throughout cerebral white matter . The latter observations suggest that diffusion of extracellular hemoglobin and ultimately iron is widespread after IVH. The mechanistic implications are clear and likely involve the free radical–mediated disturbances described in experimental models, as described earlier.

The disturbances of pre-OLs and the presence of diffuse white matter gliosis observed in the presence of severe IVH are similar to those observed in premature infants with typical cerebral white matter injury without IVH (see Chapters 18 and 19 ). Indeed, classic cystic periventricular leukomalacia, the most severe form of white matter injury, is more common in infants with severe IVH than in those without IVH. Although difficult to distinguish in vivo, it is likely that most infants with severe IVH and PHHC have, in addition to the abnormalities described here, degrees of cerebral white and gray matter pathology indicative of the encephalopathy of prematurity, the typical nonhemorrhagic pathology of the premature infant.

Finally, neuroimaging studies support the notion of widespread disturbances of cerebral white matter and cortical development in the presence of IVH. For example, infants with only mild degrees of IVH exhibit diminished cerebral cortical volumes and impaired cortical development with poorer neurodevelopmental outcome than do infants with no IVH. Perhaps most striking, recent advanced MRI studies in infants with IVH have shown a striking relation between apparent cerebral white matter dysmaturation and disturbed resting state functional connectivity ( Fig. 28.18 ).

The severity of white matter injury was also shown to correspond with reductions in functional connectivity. Specifically, reduced connectivity between the motor cortex and thalamus was proportional to the degree of white matter injury, and infants with high-grade IVH from the same cohort were found to have impaired structural connectivity in the posterior limb of the internal capsule, corpus callosum, and optic radiations.

Elevations of Arterial Blood Pressure and Pressure-Passive Cerebral Circulation

Concerning the role of elevations in arterial blood pressure, the presence of a pressure-passive cerebral circulation would be expected to lead to an increase in cerebral blood flow in association with increases in blood pressure, with the potential consequence being rupture of vulnerable germinal matrix vessels. The striking increase in cerebral blood flow associated with increases in blood pressure can be shown in real time by NIRS (see Fig. 28.21 ). A decisive demonstration of the relation between pressure-passive cerebral circulation and the occurrence of IVH was obtained from a classic study of 57 preterm infants supported by mechanical ventilation during at least the first 48 hours of life ( Fig. 28.22 ). Infants in whom ultrasonographic signs of severe IVH developed had prior evidence of a pressure-passive cerebral circulation, whereas those with intact cerebrovascular autoregulation developed either no hemorrhage or only mild hemorrhage ( Fig. 28.23 ). The work of Tsuji and co-workers showed that 47% of infants with impaired cerebrovascular autoregulation developed IVH (or periventricular leukomalacia, or both), whereas only 13% of those with intact autoregulation developed these lesions. Consistent with a potential role for arterial hypertension in this setting is the demonstration of a relationship between maximum systolic blood pressure above a threshold value and subsequent occurrence of IVH. The limit for the highest tolerable peak systolic blood pressure was markedly lower for the smaller infants. A particular role for minute-to-minute alterations in blood pressure has also been demonstrated.

A more recent study continued to suggest the importance of elevations in cerebral perfusion in causation of high-grade IVH. Thus cranial Doppler studies for middle cerebral artery cerebral blood flow velocity in 185 preterm infants who were receiving mechanical ventilation showed that severe IVH (grades III to IV) was associated with an elevation in diastolic closing margin measure of cerebral perfusion in diastole that exceeds “cerebral closing margin.” The measures were a complex combination of assumptions based on Doppler-based estimations of cerebrovascular resistance and compliance. This modeling requires replication, but the findings suggested that high-grade IVH was associated with excessive cerebral perfusion. The timing of this elevation in relation to the timing of the IVH was not delineated within the study. However, although this and other studies have shown increased crSO ₂ in the first few days of life to be associated with GM/IVH, others have found that GM/IVH was associated with decreased crSO ₂ immediately after birth and within the first few days after birth. A recent study of a cohort of extremely preterm newborns monitored by NIRS for exploratory analysis showed that infants with no GM/IVH had relatively stable crSO ₂ in the first 108 hours after birth, whereas those who developed GM/IVH showed a decline in their crSO ₂ after the second day of life. The mechanism of the association of GM/IVH with low crSO ₂ was not well defined, and it remained unknown whether the association between low crSO ₂ and GM/IVH was causal. Thus the exact sequence of changes in cerebral perfusion, reflected by cerebral saturation on NIRS, remains under investigation and may be assisted by correlative targeted echocardiography and invasive blood pressure monitoring. In summary, hypoperfusion of the preterm brain, which is often followed by hyperperfusion or fluctuations in cerebral perfusion, is associated with an elevated risk for IVH . The ability to fully capture the nature and timing of the cerebral hypoperfusion by cererbal saturation monitoring with NIRS technology is unclear. Low cerebral saturations may also be later reflecting the presence of established brain injury with a reduced cerebral oxygen saturation. Thus advances in technology to better reflect cerebral perfusion may improve insights, especially when linked with measures of oxygen extraction or utility.

Moreover, as discussed in Chapters 16 and 19 , the upper limit of the normal autoregulatory range in the infant is dangerously close to the upper limit of the range of normal blood pressure. Studies in developing animals indicate that the receptor number for specific vasoconstricting prostaglandins, which are important in setting the upper limit of the autoregulatory range in the adult, are low early in maturation and thereby impair protection of the cerebral circulation from increases in blood pressure.

Whether the pressure-passive cerebral circulatory state relates to dysfunctional autoregulation per se, to maximal vasodilation caused by hypercarbia or hypoxemia (or both), to the cranial trauma of even a normal vaginal delivery, to dopamine therapy for hypotension, or to normal arterial blood pressures in the premature infant that are dangerously close to the upslope of a normal autoregulatory curve remains unclear. Experimental support for these several possibilities is available (see Chapter 16 ). Whatever the mechanism, however, the balance of current data imparts particular importance to events that cause elevations in arterial blood pressure, especially abrupt elevations, in the small premature infant.

Causes of Increased Arterial Blood Pressure in the Human Newborn

The causes of abrupt elevations in arterial blood pressure sometimes shown to be accompanied by increased cerebral blood flow velocity by the Doppler technique, or increased cerebral blood volume by NIRS in the premature infant, are clearly important to detect (and to prevent, whenever possible) ( Table 28.6 ). These causes include the following: physiological events such as rapid eye movement sleep and the first minutes and hours after birth; caretaking concomitants such as inadvertent noxious stimulation, abdominal examination, handling (see Figs. 28.11 and 28.24 ), instillation of mydriatics, and tracheal suctioning ( Fig. 28.25 ); systemic complications such as pneumothorax and rapid infusion of colloid; and neurological complications such as seizures. Avoidance of such handling and rapid interventions in the very preterm infant is included in many care bundles aimed at reducing GMH-IVH.

Although the degree to which these events contribute to the pathogenesis of IVH requires further quantitation and probably depends on concomitant clinical circumstances, particular importance can be attributed to pneumothorax . In one earlier study of nine infants, pneumothorax was accompanied consistently by abrupt elevations of systemic blood pressure and cerebral blood flow velocity, and these circulatory changes were followed within hours by IVH. Studies in newborn dogs documented abrupt increases in arterial blood pressure on rapid evacuation of pneumothorax. Thus both clinical and experimental data emphasize the potentially deleterious circulatory effects of neonatal pneumothorax.

The complexity of interaction between respiratory and cardiovascular factors in the pathway to IVH is also notable with regard to pneumothorax. A major reduction in the risk of pneumothorax occurred following the administration of exogenous surfactant therapy. Indeed, the administration of surfactant reduced the risk of pneumothoraxes by almost 50% (relative risk [RR] = 0.63; 95% CI, 0.53–0.75). However, despite this reduction in pneumothorax with surfactant administration, there has been no reduction in the incidence of IVH. One possible explanation for this lack of reduction in IVH may relate to changes in cardiovascular and respiratory stability during surfactant administration. As early as 1992, it was noted that during surfactant administration adverse changes in systemic and cerebral oxygenation could be seen. This has been replicated with less severe impact in recent years, with 62% of premature infants displaying reductions in cerebral electrophysiological activity with intubation and surfactant administration. Thus surfactant administration may have a double-edge effect with a positive effect with reduction of pneumothorax being offset by a potential negative effect of reduced cerebral perfusion during its administration. To determine the effect of therapies during this period of physiological instability in the preterm infant, monitoring of cerebral perfusion could provide much needed guidance.

Relevant Experimental Studies: Role of Hypertension

The particular importance of abrupt increases in systemic blood pressure and cerebral blood flow in pathogenesis has been demonstrated conclusively in elegant experimental studies in the newborn beagle puppy and in the preterm sheep fetus. The newborn puppy, which has been studied most extensively, has a subependymal germinal matrix approximately comparable to that of the human premature infant of 30 to 32 weeks of gestation. GMH-IVH is produced most readily in this animal by a sequence of hypotension and hypertension produced by blood removal and volume reinfusion ( Fig. 28.26 ). The marked increase in germinal matrix flow provoked by hypertension has been demonstrated strikingly by autoradiography ( Fig. 28.27 ).

Rapid Volume Expansion

The role of rapid volume expansion (see Table 28.4 ) involves not only the administration of blood or other colloid, as described in relation to systemic hypertension, but also the administration of hyperosmolar materials, such as hypertonic sodium bicarbonate. Pressure-passive cerebral circulation may not be the sole or even the principal means by which such infusions may lead to IVH, particularly in the case of sodium bicarbonate . Although the dangers of rapid infusion of hyperosmolar solutions had been noted for many years, an association of IVH in the premature infant administered sodium bicarbonate was emphasized initially by Simmons and co-workers from study of an autopsy population. The association was later confirmed in a CT study of premature infants, and the importance of rapidity of infusion was made apparent. Conflicting reports on the pathogenetic role of sodium bicarbonate relate in part to the failure to take into account such factors as rapidity of administration and also to the problems of extrapolating data to living infants from studies of dead infants, particularly in the case of IVH. At any rate, the mechanism for the effect of rapid infusion of hyperosmolar sodium bicarbonate on intracranial hemorrhage may relate in part to the abrupt elevation of arterial pressure of carbon dioxide (Pa co ₂ ) that results in the poorly ventilated or nonventilated patient from the buffering effect of the bicarbonate. The elevated Pa co ₂ would then act on cerebral arterioles, by causing an increase in perivascular hydrogen ion (H ⁺ ) concentration, to increase cerebral perfusion as outlined next. Finally, related to the issue of rapid volume expansion, the data related to umbilical cord milking support its risk as a source of rapid intravascular volume expansion. In the most recent retrospective large observational study from the National Institute of Child Health neonatal network, data were collected on 1834 infants, 23.6% of whom were exposed to umbilical cord milking and 76.4% of whom were exposed to delayed cord clamping. Infants exposed to umbilical cord milking had higher odds of severe IVH (19.8% umbilical cord milking vs. 11.8% delayed cord clamping; adjusted odds ratio [aOR] = 1.70; 95% CI, 1.20–2.43). Other outcomes were similar between groups.

Hypercarbia

The role of hypercarbia in causing increases in cerebral blood flow of pathogenetic importance for IVH may be appreciable in selected infants. Hypercarbia, a common accompaniment of respiratory distress syndrome, respiratory complications, apneic episodes, and so forth, has been conclusively demonstrated to be a potent means for increasing cerebral blood flow in experimental studies (see Chapter 16 ). Indeed, careful studies of mechanically ventilated preterm infants show a pronounced reactivity of cerebral blood flow to changes in Pa co ₂ (≈30% increase in cerebral blood flow per kilopascal increase in Pa co ₂ ) after the first 24 hours of life. Notably, in the first 24 hours of life, this normal reactivity was attenuated markedly (≈10% increase in cerebral blood flow per kilopascal increase in Pa co ₂ ) in mechanically ventilated infants with normal subsequent ultrasonograms, but it was actually absent in infants with subsequent severe IVH. This observation suggested that, in the first day of life at least, hypercarbia of at least a moderate degree may not be a major pathogenetic factor for severe IVH in mechanically ventilated infants. A similar lack of correlation between hypercarbia and IVH was apparent in several other studies. An increased risk for IVH after hypercarbia, however, was suggested in several other reports, including three that employed multivariate analysis. In a particularly large study ( n = 463), hypercarbia (defined as Pa co ₂ >60 mm Hg) showed a positive relation with IVH. In a later study of permissive hypercapnia to 45 to 55 mm Hg (vs. 35–45 mm Hg in the control group) in ventilated premature infants, no statistically significant difference in IVH was noted between the groups, although the incidence of severe IVH was 29% in the permissive hypercapnia group versus 20% in the control group (not statistically significant). Thus a role for hypercarbia in pathogenesis of IVH may require particularly marked elevations of Pa co ₂ . Consistent with this speculation is the demonstration that hypercapnia leads to clearly impaired autoregulation at Pa co ₂ levels above 45 mm Hg ( Fig. 28.28 ). Such levels were shown to be significantly associated with the occurrence of severe IVH and periventricular hemorrhagic infarction in a study of 58 infants. In the most recent study of the role of P co ₂ in brain injury and outcomes in the preterm infant, an association was found between the severity and length of time of hypercarbia and IVH and adverse outcome. However, the authors hypothesized that the hypercarbia was only a biomarker for the underlying severity of respiratory disease that may be the greater mediator for adverse neurodevelopmental outcomes. However, the majority of data support that hypercarbia and rapid fluxes in CO ₂ result in rapid fluxes in cerebral perfusion and increased risk for IVH.

Decreased Hemoglobin

The role of decreased hematocrit in causing increases in cerebral blood flow of pathogenetic importance for IVH may be greater than was previously suspected. Thus, as described in Chapter 16 , an inverse correlation exists in the human infant between hemoglobin concentration and cerebral blood flow as well as between the concentration of adult versus fetal hemoglobin (higher hemoglobin oxygen affinity) and cerebral blood flow. In one study of premature infants in the first days of life, cerebral blood flow increased by 12% per 1-mM decrease in hemoglobin. The inverse relationship between hematocrit and cerebral blood flow described previously in experimental studies has been suggested to result from changes in arterial oxygen content or blood viscosity. Because alterations in newborn hematocrit to less than 60% have little influence on blood viscosity, the major factor in the studies of human infants is considered to be related to arterial oxygen content and thereby cerebral oxygen delivery. Cerebral blood flow presumably increases to maintain cerebral oxygen delivery at a constant level. Consistent with this possibility, apparently stable premature infants with low hematocrits (<21%) had clinically unsuspected high cardiac output. The adaptive response of increased cerebral blood flow may become maladaptive if certain vulnerable capillary beds (e.g., in the germinal matrix) are exposed to the elevated cerebral blood flow. When one considers that iatrogenic blood loss, owing to repeated blood sampling, and low initial blood volume are common in sick premature infants, especially during the periods of highest risk for occurrence of IVH, the role of decreased hematocrit as a cause of IVH could be considerable.

Potentially consistent with supporting the role of anemia in the risk for IVH, a recent study demonstrated an increased risk of severe IVH among VLBW infants following a red cell transfusion within the first 72 hours of birth. This finding remained significant after controlling for confounding variables (RR = 2.02; 95% CI, 1.54–3.33). The authors were unable to determine the underlying mechanism given the retrospective nature of the review, but one possibility for the increased risk is concomitant anemia requiring transfusion, rather than the transfusion itself. However, the infants developing IVH may have been sicker and thus at greater risk. Indeed, although there was no difference in coagulation measures, the infants with IVH received more frozen plasma and platelet transfusions and had longer ampicillin courses, higher nucleated red blood cell counts, more vasopressor use, and a higher mortality rate.

Relevant in this context are results of studies of the timing of cord clamping. Delayed cord clamping (DCC; 30 seconds) compared with immediate cord clamping (6–7 seconds) is associated with a slightly higher baseline hematocrit (49% vs. 46%), higher mean blood pressure (33.8 vs. 31.9 mm Hg), increased superior vena caval flow, and a significantly reduced risk for IVH (14% vs. 36%). This reduction in IVH with DCC was supported by a meta-analysis of the randomized trials in preterm infants, although it is noteworthy that the largest studies included larger preterm infants with birthweights >1500 g. Because of these benefits in preterm infants and the advantages of increasing iron stores in healthy term-born infants, the American College of Obstetricians and Gynecologists recommended a delay in umbilical cord clamping in vigorous term and preterm infants for at least 30 to 60 seconds after birth. The most recent Cochrane review of delayed cord clamping, including 29 studies, concluded that there was a reduction in death or disability (RR = 0.61; 95% CI, 0.39–0.96), with no clear evidence of reduction in IVH.

Blood Glucose

Decreased blood glucose now should be considered in the evaluation of pathogenetic factors for IVH in view of the observation that cerebral blood flow increases twofold to threefold when blood glucose declines to levels lower than 1.7 mM in the premature infant. Blood glucose levels lower than 1.7 mM in premature infants are not unusual in the first days of life in many neonatal intensive care units (see Chapter 29 ).

Increased blood glucose has also been evaluated as a potential risk factor for IVH. In a recent case-control study of high-grade IVH ( n = 70) compared with no IVH ( n = 108), infants with IVH had significantly more hyperglycemic events (2.9 ± 1.7 vs. 2.4 ± 1.8 events, p < .05) with longer duration (22.2 ± 14.2 vs. 14.1 ± 12.5 hours, p < .001) and a higher hyperglycemic index (1.0 ± 0.9 vs. 1.4 ± 1.0, p = .003). Respiratory distress syndrome, hypotension, and thrombocytopenia increased the aOR for IVH. Hypoglycemia was not independently associated with IVH. Conversely, the increase in hyperglycemic duration most prominently increased the aOR for severe IVH (OR = 10.33; 95% CI, 10.0–10.6; p = .033). To avoid hyperglycemia, insulin therapy is often initiated. However, an important randomized controlled trial of tight glycemic control with insulin versus standard care documented a nonsignificant trend toward an increase in the incidence of grade III/IV IVH in the insulin-treated group (insulin 6/38 infants, 14%, vs. standard care 3/43, 7%, p = .35). The insulin-treated group did have more episodes of hypoglycemia. Thus avoidance of protracted hyperglycemia and hypoglycemia may be most prudent as further data are collected on this clinical factor.

Finally, it has been suggested that alterations in the osmotic gradient may occur with hyperglycemia and other metabolic derangements, such as hypernatremia, leading to an increase in the intravascular pressure relative to the surrounding extravascular tissue that may predispose to IVH. Several cohort studies have shown that those states associated with an alteration in the osmotic balance, such as hyperglycemia and hypernatremia (even high sodium intake in the absence of hypernatremia), are associated with an increased risk for IVH. However, the retrospective nature of these studies cannot delineate the underlying mechanism for this increased risk.

Importance of Venous Anatomy

The particular importance of increased venous pressure in pathogenesis of IVH relates in part to the venous anatomy in the region of the germinal matrix (see Fig. 28.7 ). Thus the direction of deep venous flow takes a peculiar U-turn in the subependymal region at the level of the foramen of Monro (i.e., the most common site of GMH). Also at this site is the point of confluence of the medullary, thalamostriate, and choroidal veins to form, in sequence, the terminal vein and then the internal cerebral vein, which ultimately empties into the vein of Galen. A recent study aimed to compare the subependymal vein anatomy of preterm infants with IVH, as evaluated by susceptibility-weighted imaging (SWI) venography, with a group of age-matched controls with normal brain MRI, to explore the relationship between the anatomical features of subependymal veins and IVH. SWI venographies of 48 infants with GMH-IVH and 130 infants with normal brain MRI were evaluated retrospectively. Subependymal vein anatomy was classified into six different patterns: type 1 represented the classic pattern and types 2 to 6 were considered anatomical variants ( Fig. 28.29 ). A significant difference was noticed among the six anatomical patterns according to the presence of IVH ( p = .014). Anatomical variants were observed with higher frequency in infants with IVH than in controls (62.2% and 49.6%, respectively). Neonates with GMH-IVH presented a narrower curvature of the terminal portion of subependymal veins ( P < .05). These anatomical features were significantly associated with IVH ( P < .05).

Labor and Delivery

Concerning labor and delivery, marked increases in cerebral venous pressure must be common accompaniments. Indeed, in one study of 46 infants, when measurement of “fetal head compression pressure” was determined by a compression transducer positioned between the fetal head and the wall of the uterus, the overall mean pressure was 158 mm Hg. Deformations of the particularly compliant premature skull are likely to accentuate the increases in venous pressure caused by normal labor. Indeed, the deleterious effects of labor (see later discussion) appear to be most pronounced in the most premature infants. The skull deformations can lead to obstruction of major venous sinuses and presumably increased venous pressure. Support for this notion has been provided by studies of blood flow velocity in the sagittal sinus, cerebral blood volume, and intracranial pressure during manipulations such as external pressure on the skull or rotation of the neck. These effects may be expected to be greater with breech delivery. Available data are somewhat inconsistent concerning a relationship between factors such as presence or absence of labor, duration of labor, mode of delivery, and the occurrence of IVH, although in general the studies were not designed to address these issues specifically and were retrospective. The inconsistency of the data, however, does not rule out a contributory role of intrapartum events in causation of IVH in certain infants. Thus in a study that addressed the role of presence or absence of labor, duration of labor, mode of labor, and potential confounders in a multivariate analysis, Leviton and co-workers showed that infants delivered vaginally were more likely to develop IVH than those delivered abdominally, that labor longer than 12 hours increased risk of IVH regardless of the mode of delivery, and that the occurrence of labor before abdominal delivery increased the incidence of IVH by two to four times, depending on the duration of labor ( Table 28.7 ). In a separate study of 201 VLBW infants, multivariate analysis also indicated an increased risk (2.2-fold) of IVH for infants delivered vaginally, a very low risk (7%) for infants delivered abdominally with no labor, and an increased risk among infants delivered abdominally for labor greater than 10 hours in duration (40%). Subsequent investigations of 229 and 254 infants, respectively, showed an increased risk of IVH occurring in the first 3 to 12 hours of life as a function of active labor and vaginal delivery. Finally, a multicenter study of 4795 infants of less than 1500 g birthweight showed an incidence of grade III and IV IVH in 19% of vaginally delivered infants and 11% of those delivered by cesarean section without labor. On balance, these data suggest that labor and delivery influence the risk of IVH in premature infants and have implications concerning a potential role for cesarean section in prevention (see later discussion).

The most recent studies of mode of delivery relative to IVH have continued to demonstrate conflicting results, with several finding no association with method of delivery. A recent study of 158 infants born at less than 1500 g found that there was an increased risk of mild IVH among infants with vaginal delivery versus cesarean section prior to the second stage of labor. The studies that did not report an association between mode of delivery and IVH did not comment on the duration of labor or the stage during which the cesarean section was performed, possibly explaining some of the discrepancies in the literature. Of note, it has been recently shown that head position, both before delivery and in the neonatal nursery, may also alter cerebral venous drainage. One study with NIRS showed that cerebral venous drainage may be impaired in prone or side positions. However, a more recent study of head position did not confirm any effect of changing head position on cerebral measures of oxygenation. The Cochrane review of the relationship of head position to neonatal outcomes in the very preterm infant reported a meta-analysis of three trials (290 infants). The review did not show an effect of midline head position on rates of GM-IVH (RR = 1.11; 95% CI, 0.78–1.56) and severe IVH (RR = 0.71; 95% CI, 0.37–1.33). The certainty of the evidence was acknowledged as very low because of limitations in study design and imprecision of estimates.

Hypoxic-Ischemic Injury

With perinatal asphyxial events, circulatory collapse may lead to hypoxic-ischemic cardiac failure and, as a consequence, increased cerebral venous pressure. The cardiac disturbance is caused by injury of papillary muscle, subendocardial tissue, and myocardium. The importance of increased venous pressure in association with asphyxia in the causation of IVH was shown in experimental studies of preterm fetal sheep. Thus it seems likely that increased venous pressure could contribute to the propensity to IVH observed after serious asphyxia. Consistent with this notion are the strong relationships among factors such as severe umbilical cord acidemia, low Apgar scores, the need for neonatal resuscitation, and the occurrence of severe IVH (see later). Other factors associated with asphyxia, such as ischemic injury to the germinal matrix and hypercarbia, are also likely important.

Respiratory Disturbances

Concerning respiratory disturbances , available data suggest that factors such as positive-pressure ventilation with relatively high peak inflation pressure, tracheal suctioning, abnormalities of the mechanics of respiration, and pneumothorax may be major causes of increased cerebral venous pressure in the premature infant. Thus, extending earlier observations, Cowan and Thoresen used Doppler measurements of blood flow velocity in the superior sagittal sinus to demonstrate a striking sensitivity of the venous circulation to the level of peak inflation pressure; the smallest infants exhibited the most marked effects.

The possibility of a particular importance for venous abnormalities in causation of IVH was raised by a study of intubated preterm infants with respiratory distress syndrome under conditions in which dangerous alterations in arterial blood pressure occur (i.e., elevations with tracheal suctioning and fluctuations with breathing out of synchrony with the ventilator ). The effects on the venous circulation were dramatic. With elevations in arterial blood pressure produced by tracheal suctioning, pronounced changes in venous pressure also occurred ( Fig. 28.30A ). Moreover, because the magnitude and the direction of the changes in venous pressure often were not similar to those in arterial pressure, striking changes in perfusion pressure resulted ( Fig. 28.30B ). Similarly, because fluctuations in arterial blood pressure were associated with noncoordinate fluctuations in venous pressure ( Fig. 28.31A ), pronounced and continuous alterations in perfusion pressure resulted ( Fig. 28.31B ). Thus under both circumstances, decreases in perfusion pressure by as much as 10 to 20 mm Hg were followed in seconds by abrupt, similar increases in perfusion pressure. Because these changes occur essentially on a beat-to-beat basis, it is unlikely that autoregulation, even if functional, could protect critical capillary beds by causing the changes in arteriolar diameter necessary to maintain constant cerebral blood flow under such circumstances. Thus the previously established role for disturbed mechanics of respiration with fluctuations in arterial blood pressure in the causation of IVH (see Table 28.5 ) may be mediated as much by alterations on the venous side of the cerebral circulation as by alterations on the arterial side. A similar conclusion can be drawn for the previously established role in causation of IVH of abruptly increased arterial blood pressure with pneumothorax because this respiratory complication has been shown to cause abruptly increased venous pressure as well. A study of 58 cases of severe IVH with periventricular hemorrhagic infarction showed a significant relationship of pneumothorax with the occurrence of the lesion.

Importance of Pressure-Passive Cerebral Circulation

Decreases in cerebral blood flow, occurring either prenatally (perhaps primarily intrapartum) or postnatally, may play an important role in pathogenesis of IVH in certain infants. The principal consequence of the decreased cerebral blood flow is injury of germinal matrix vessels, which rupture subsequently on reperfusion. The importance of vascular border zones and end zones in the matrix, as well as the intrinsic vulnerability of the matrix vessels to oxygen deprivation, is emphasized later (see section on vascular factors). As indicated earlier, hemorrhagic hypotension preceding volume reexpansion is the optimal means to produce IVH experimentally in the newborn beagle puppy. In the premature infant, decreases in cerebral blood flow are most likely with perinatal hypoxia-ischemia and with various postnatal events that result in systemic hypotension. Because of the pressure-passive cerebral circulation in sick premature infants, this hypotension can lead to a decrease in cerebral blood flow. Recall that a detailed study of 90 premature infants in the first 5 days of life showed that more than 95% had pressure-passive periods, with a mean total time of pressure-passivity of 20%.

Perinatal Hypoxic-Ischemic Events

Although it is not an obligatory event for development of IVH, the infant with prior perinatal asphyxia clearly has an increased likelihood of developing IVH and, in our experience, the hemorrhage in such infants tends to be relatively large. Indeed, a study of 58 infants with periventricular hemorrhagic infarction found a strong association of the lesion with fetal distress and the need for emergency cesarean section, low Apgar scores, and the need for respiratory resuscitation. Perinatal hypoxic-ischemic events presumably explain, at least in part, the relation between low Apgar scores, early acidosis, early use of bicarbonate or pressors, hypocarbia, and the subsequent overall occurrence of IVH, particularly lesions that develop in the first 12 hours. The mechanism for provocation of IVH with perinatal asphyxia is complex and includes increases in cerebral blood flow associated with impaired vascular autoregulation, increases in cerebral venous pressure, and decreases in cerebral blood flow associated with hypotension, with resulting injury to matrix capillaries. Release of endogenous vasodilators (e.g., adrenomedullin) in the first hours after birth may also play a role in this situation. Studies of the concentrations of brain-specific creatine kinase isoenzymes in cord blood or blood samples obtained early in the postnatal period or of the hypoxanthine metabolite uric acid in blood samples obtained on the first postnatal day, in preterm infants who later developed IVH, support the notion that late intrauterine injury (perhaps asphyxia) may be involved in at least some cases of IVH. In one study, those infants who developed IVH had cord blood levels of brain-specific creatine kinase that were six times greater than those in the infants who did not develop IVH. Moreover, the possibility of a perinatal hypoxic-ischemic insult to the brain as a predecessor to IVH was suggested by the finding of depressed amplitude-integrated electroencephalographic activity before the occurrence of the hemorrhage in 4 of 10 carefully studied preterm infants. In a separate study, early continuous electroencephalographic monitoring detected abnormalities such as excessive discontinuity before the occurrence of IVH. These abnormalities are similar to those produced by hypoxic-ischemic insults (see Chapter 13 ).

More recent studies have documented a high incidence of seizures in the preterm infant in the first 72 hours of life, accompanied by a strong association with the presence of IVH. In one study, 95 VPT infants underwent amplitude-integrated electroencephalography monitoring during the first 72 hours of life. The overall incidence of seizures in this sample was 48%. High seizure burden was associated with increased risk of IVH throughout each of the 3 days of monitoring. The seizures observed in this very preterm cohort demonstrated a similar evolution in time line to the seizures of a term infant suffering from hypoxic-ischemic encephalopathy with the highest median seizure burden during the 0- to 24-hour period. These observations support a potential role of perinatal asphyxia in the pathway to IVH.